Transcriptional signatures modulating shoot apical meristemmorphometric and plant architectural traits enhance yieldand productivity in chickpea
Laxmi Narnoliya1, Udita Basu1, Deepak Bajaj1, Naveen Malik1, Virevol Thakro1, Anurag Daware1, Akash Sharma1,
Shailesh Tripathi2, Venkatraman S. Hegde2, Hari D. Upadhyaya3, Ashok K. Singh2, Akhilesh K. Tyagi1,4 and
Swarup K. Parida1,*1Genomics-assisted Breeding and Crop Improvement Laboratory, National Institute of Plant Genome Research (NIPGR),
Aruna Asaf Ali Marg, New Delhi 110067, India,2Division of Genetics, Indian Agricultural Research Institute (IARI), New Delhi 110012, India,3International Crops Research Institute for the SemiArid Tropics (ICRISAT), Patancheru, Telangana 502324, India, and4Department of Plant Molecular Biology, University of Delhi, South Campus, New Delhi 110021, India
Received 11 September 2018; revised 31 January 2019; accepted 7 February 2019; published online 11 February 2019.
*For correspondence (e-mails [email protected]; [email protected]).
SUMMARY
Plant height (PH) and plant width (PW), two of the major plant architectural traits determining the yield and
productivity of a crop, are defined by diverse morphometric characteristics of the shoot apical meristem
(SAM). The identification of potential molecular tags from a single gene that simultaneously modulates
these plant/SAM architectural traits is therefore prerequisite to achieve enhanced yield and productivity in
crop plants, including chickpea. Large-scale multienvironment phenotyping of the association panel and
mapping population have ascertained the efficacy of three vital SAM morphometric trait parameters, SAM
width, SAM height and SAM area, as key indicators to unravel the genetic basis of the wide PW and PH trait
variations observed in desi chickpea. This study integrated a genome-wide association study (GWAS); quan-
titative trait locus (QTL)/fine-mapping and map-based cloning with molecular haplotyping; transcript profil-
ing; and protein-DNA interaction assays for the dissection of plant architectural traits in chickpea. These
exertions delineated natural alleles and superior haplotypes from a CabHLH121 transcription factor (TF)
gene within the major QTL governing PW, PH and SAM morphometric traits. A genome-wide protein-DNA
interaction assay assured the direct binding of a known stem cell master regulator, CaWUS, to the
WOX-homeodomain TF binding sites of a CabHLH121 gene and its constituted haplotypes. The differential
expression of CaWUS and transcriptional regulation of its target CabHLH121 gene/haplotypes were appar-
ent, suggesting their collective role in altering SAM morphometric characteristics and plant architectural
traits in the contrasting near isogenic lines (NILs). The NILs introgressed with a superior haplotype of a
CabHLH121 exhibited optimal PW and desirable PH as well as enhanced yield and productivity without com-
promising any component of agronomic performance. These molecular signatures of the CabHLH121 TF
gene have the potential to regulate both PW and PH traits through the modulation of proliferation, differen-
tiation and maintenance of the meristematic stem cell population in the SAM; therefore, these signatures
will be useful in the translational genomic study of chickpea genetic enhancement. The restructured culti-
vars with desirable PH (semidwarf) and PW will ensure maximal planting density in a specified cultivable
field area, thereby enhancing the overall yield and productivity of chickpea. This can essentially facilitate the
achievement of better remunerative outputs by farmers with rational land use, therefore ensuring global
food security in the present scenario of an increasing population density and shrinking per capita land area.
Keywords: chickpea, desi, fine mapping, genome-wide association study, kabuli, map-based cloning, near
isogenic lines, plant height, plant width, productivity, quantitative trait locus, RIL, shoot apical meristem,
single nucleotide polymorphisms, transcription factor, yield.
© 2019 The AuthorsThe Plant Journal © 2019 John Wiley & Sons Ltd
864
The Plant Journal (2019) 98, 864–883 doi: 10.1111/tpj.14284
INTRODUCTION
Plant architecture is the three-dimensional organization of
the plant body and is represented by various major com-
ponent traits such as the inflorescence architecture, plant
height (PH), plant width (PW), branching pattern, and
others. The plant architecture determines the ultimate
agronomic value of a crop by influencing its photosyn-
thetic efficiency and maturity time, and therefore overall
crop growth, development and production. In the present
scenario of unforeseen climate change and diminishing
cultivable land area, the development and use of proficient
strategies for enhancing crop yield is essential to ensure
global food security. Ever since the advent of the Green
Revolution in the 1940s, the manipulation of plant architec-
ture has clearly arisen as an efficient strategy to enhance
crop yield and productivity within land and climatic con-
straints. In this regard, significant accomplishments have
been achieved with the development of lodging-resistant
semidwarf varieties of wheat and rice by introducing the
genes/alleles that control one of the vital plant architecture
traits, PH, leading to enhanced grain yield in these staple
food crops (Peng et al., 1999; Evenson and Gollin, 2003).
An ideal plant architecture also contributes to the adapta-
tion of crops towards different environmental stress along
with enhancing their photosynthetic efficiency, yield, and
harvest index (Reinhardt and Kuhlemeier, 2002; Sakamoto
et al., 2004; Yang and Hwa, 2008). Plant width is another
vital plant architectural trait that has a direct impact on the
efficient use of land area under cultivation. With limiting
land resources, optimizing PW without compromising yield
per plant will accommodate a greater number of plants in
a specified cultivable land area and thereby significantly
enhance the crop yield and productivity.
Manipulation of plant architecture requires a compre-
hensive understanding of the genetic networks and regula-
tory pathways governing its various major contributing
traits in crops (Pickersgill, 2007; Jin et al., 2008). Decipher-
ing the genetic and molecular mechanism that underlies
plant architecture variation will not only address the basic
fundamental intricacies but also facilitate genomics-
assisted breeding to develop high-yielding cultivars
restructured with an ideal plant type (Wang and Li, 2008).
The gene regulatory networks within the meristematic cells
play a major role in defining the plant architecture. The
shoot apical meristem (SAM) is the key organizing centre
of stem cells and controls most developmental traits con-
tributing to seed yield in crop plants. It comprises a dome
of pluripotent cells that ultimately generates all the organs
of the plant shoot and performs the dual functions of
organogenesis and stem cell maintenance (Barton, 2010;
Thompson et al., 2014, 2015). SAM morphology, a quanti-
tative trait defined by the descriptive measurements of
SAM shape and size, can be exploited as target traits for
the genetic enhancement of crop plants. Henceforth, the
phenotypic variations of said SAM morphometric charac-
teristics are correlated with allelic diversity scanned from
natural and mapping populations through genetic and
association mapping to successfully identify potential
genomic loci governing diverse plant architectural traits in
major food crops (Thompson et al., 2014, 2015; Leiboff
et al., 2015; Cai et al., 2016; Saxena et al., 2017). These
studies evidently suggest that comprehensive genetic dis-
section of SAM morphometric and ideal plant architectural
traits is vital for crop improvement. This will essentially
therefore lay the foundation for designing efficient geno-
mics-assisted breeding strategies to improve yield and pro-
ductivity in crop plants.
Chickpea (Cicer arietinum L.) is a vital nutrient-rich diploid
legume crop with a genome size of ~740 Mbp (Kumar et al.,
2011; Varshney et al., 2013a). The genome sequences of cul-
tivated desi and kabuli chickpea and their close wild progeni-
tor C. reticulatum have been decoded in the recent past
(Jain et al., 2013; Varshney et al., 2013b; Parween et al.,
2015; Gupta et al., 2016). These reference genomic resources
efficiently facilitate discovery and high-throughput genotyp-
ing of sequence-based markers including simple sequence
repeats (SSRs) and single nucleotide polymorphisms (SNPs)
among the large-scale germplasm accessions (association
panel) and mapping populations of chickpea. These exer-
tions drive genome-wide association studies (GWAS) and
quantitative trait locus (QTL) mapping that detected several
major QTLs and genes governing multiple yield-contributing
traits in chickpea (Saxena et al., 2014a; Bajaj et al., 2015a,b;
Kujur et al., 2015a,b, 2016; Upadhyaya et al., 2015). Most of
these genomics-assisted breeding efforts employed to date
in trait dissection for chickpea genetic enhancement are
intended to improve the yield per plant. Unfortunately,
genetic manipulation of plant architectural traits has yet not
been explored for genetic enhancement of this crop legume
to increase its yield and productivity.
Chickpea plants are well known for their non-synchro-
nized spreading to semispreading/semierect growth habits,
which usually increase the PW and therefore require more
space for suitable growth and development. As a conse-
quence, fewer plants can be accommodated in a definite
cultivable land area, resulting in low chickpea productivity.
In this context, the optimization of PW in addition to PH is
essential for the development of lodging-resistant semid-
warf plant type chickpea cultivars. Restructured plants with
an ideal architecture will enhance the yield and productiv-
ity of chickpea and therefore can be useful in providing
higher remuneration, especially to small and marginal
farmers with minimal land resources. Delineation of func-
tionally relevant genes and alleles governing both SAM
morphometric characteristics as well as plant architectural
© 2019 The AuthorsThe Plant Journal © 2019 John Wiley & Sons Ltd, The Plant Journal, (2019), 98, 864–883
Transcriptional tags enhance chickpea productivity 865
traits (PH and PW) using diverse integrated genomic strat-
egy is therefore essential. These delineated molecular tags
can further be deployed in marker-aided selection for
restructuring desirable new plant types, which will be use-
ful to enhance yield and productivity in chickpea.
Keeping that in mind, the current investigation inte-
grated diverse molecular genetics approaches with geno-
mics-assisted breeding and functional genomic strategies
to decode the complex genetic architecture of vital SAM
morphometric trait characteristics as well as two major
plant architectural traits in chickpea, PW and PH. The
GWAS, regional association analysis, QTL/fine-mapping
and map-based cloning were correlated with gene haplo-
type-specific association and transcript profiling and the
protein�DNA interaction assay in the constituted associa-
tion panel (desi and kabuli germplasm accessions), as well
as recombinant inbred lines (RIL) and NIL mapping popula-
tions of chickpea with contrasting plant and SAM architec-
tural traits. This result assisted us in delineating the most-
promising molecular signatures (genes, natural alleles, and
superior haplotypes) regulating vital plant architectural
traits, PW and PH through the modulation of SAM morpho-
metric features, with an aim to drive high yield and pro-
ductivity in chickpea.
RESULTS
Phenotypic evaluation and characterization of an
association panel for major plant architectural and SAM
morphometric traits in chickpea
An association panel comprising 291, including 189 desi
and 102 kabuli, germplasm accessions was employed for
genetic dissection of major plant and SAM architectural
traits through GWAS in chickpea (Table S1). The multienvi-
ronment (field and controlled conditions) replicated pheno-
typic evaluation and characterization of the association
panel were performed for two vital plant architectural
traits, PW and PH, and seven SAM morphometric trait
parameters, including SAM width (SWi), SAM height (SH),
SAM midpoint width (SMWi), SAM height/width ratio (SH/
SWi), SAM arc length (SAL), SAM area (SA) and SAM
radius (SR), based on histology (Figure 1). A comprehen-
sive desi and kabuli cultivar-wise analysis exhibited the
wider level of phenotypic variation for five of these major
architectural traits (PW, PH, SWi, SH, and SA) among desi
accessions (6.08–23.35% coefficient of variation (CV)) com-
pared with that among kabuli accessions (1.69–9.27% CV).
A greater degree of phenotypic variation was observed for
the SH (23.35% CV) followed by the SA (19.84%) and PH
(11.74%) traits among 189 desi accessions (Table S2 and
Figure 2a, b). In kabuli, phenotypic variation was maxi-
mum for the PH (9.03% CV) followed by the PW (6.38%)
and SA (3.20%) traits among 102 accessions. The normal
frequency distribution was evident, particularly for the PW
and PH traits phenotyped in the multienvironment field
conditions among 291 desi and kabuli accessions (Fig-
ure 2a). The high broad-sense heritability (H2) of the PW,
PH, SWi, SH, and SA traits (75–83%) among the desi and
kabuli accessions was apparent. A detailed correlation
analysis and subsequent hierarchical cluster display of the
studied SAM morphometric and plant architectural traits
were performed among 291 desi and kabuli accessions. A
significant positive correlation (65–73%) among PW, PH,
SWi, and SH traits in all these desi and kabuli accessions
was evident (Figures 1d and 2c). Comparatively, a high
correlation (91–93%) among these plant architectural and
SAM morphometric traits in 189 accessions of desi chick-
pea was observed.
GWAS and regional association analysis scans of potential
genomic loci associated with major plant architectural and
SAM morphometric traits in chickpea
Genome-wide association study was performed to identify
the potential genomic loci associated with two vital plant
architectural traits (PW and PH) and three major SAM mor-
phometric traits (SWi, SH and SA) in chickpea. Primarily,
high-throughput genotyping of SNPs mined by resequenc-
ing of 291 desi and kabuli accessions representing an
association panel was carried out using a genotyping-
by-sequencing (GBS) assay. Overall, this procedure
detected 333 001 high-quality SNPs with an average fre-
quency of 625.6 SNPs per Mb (Table S3). Detailed structural
annotation of SNPs in diverse coding and non-coding
sequence components of the genome/genes are depicted in
Figure S1 and described briefly in Data S1. The molecular
diversity and genetic relatedness analysis based on the
unrooted phylogenetic tree, population structure and princi-
pal component analysis (PCA) differentiated all 291 acces-
sions from each other and overall grouped these accessions
into two distinct population groups, POP I and POP II (Fig-
ure S2a–c). The linkage disequilibrium (LD) estimate in an
association panel illustrated an overall LD decay (r2 reduced
to half of its maximum value) nearly at a 150-kb physical dis-
tance of the chickpea chromosomes (Figure S2d).
For GWAS, high-throughput genotyping data for the
333 001 SNPs assayed in 291 desi and kabuli accessions (as-
sociation panel) were correlated with their molecular diver-
sity, population genetic structure and PCA information, as
well as the multienvironment (field/controlled condition)
replicated phenotyping data for the PW, PH, SWi, SH, and SA
traits. GWAS at a false discovery rate (FDR) cut-off ≤0.05using multiple statistical methods of association mapping
detected six SNP loci that showed a significant association
with the PW, PH, SWi, SH, and SA traits at P ≤ 10�8 (Table 1
and Figure 3a). All six trait-associated SNPs were well vali-
dated by a general linear model (GLM) and a mixed linear
model (MLM) of TASSEL as well as a compressed mixed lin-
ear model (CMLM), population parameters previously
© 2019 The AuthorsThe Plant Journal © 2019 John Wiley & Sons Ltd, The Plant Journal, (2019), 98, 864–883
866 Laxmi Narnoliya et al.
determined (P3D) and efficient mixed model association
eXpedited (EMMAX) strategies of Genome Association and
Prediction Integrated Tool (GAPIT). Of these, the CMLM strat-
egy of GAPIT taking into account population structure and
principal components as well as ancestry were found to be
the best model for GWAS of plant architectural and SAM
morphometric traits in chickpea. These associated SNPs
were physically mapped on three chromosomes (Ca4, Ca6,
and Ca8) (Table 1 and Figure 3a). Five of six trait-associated
SNP loci derived from the diverse coding sequence compo-
nents of five genes exhibited non-synonymous amino acid
substitutions. The remaining one associated SNP was
detected in the intergenic region of the chickpea genome.
The phenotypic variation explained (PVE) for the studied
plant and SAM architectural traits by six significant SNPs in
a constituted association panel varied from 15 to 34% R2
(Table 1). Compared with five other trait-associated SNPs, a
coding SNP (G/A) showing a non-synonymous amino acid
substitution (Arginine: CGT to histidine: CAT) in a basic
helix�loop�helix (bHLH) transcription factor (TF) gene
exhibited a strong association (10�9 P with 34% combined
PVE) with the PW, PH, SWi, SH, and SA traits (Table 1 and
Figure 3a). This identified bHLH TF gene was further classi-
fied as ‘CabHLH121’ with a Myc-type functional domain, in
accordance with the comprehensive genome-wide identifica-
tion of 135 bHLH TF-encoding genes annotated from the kab-
uli chickpea genome. Therefore, a strong plant architectural
(PW and PH) and SAM morphometric (SWi, SH, and SA)
trait-associated SNP detected in a CabHLH121 TF gene
employing high-resolution GWAS was considered a promis-
ing candidate to drive genomics-assisted breeding and crop
improvement of chickpea.
Figure 1. SAM morphometric trait variation among desi and kabuli chickpea accessions.
(a) SAM architecture of a chickpea accession examined based on its seven major morphometric trait parameters (SWi, SMWi, SH, SH/SWi, SA, SAL, and SR)
using a histological assay.
(b) Histology of four desi (ICC 5590 and ICC 6306) and kabuli (ICC 12968 and ICC 16814) accessions in accordance with the indicated major morphometric trait
parameters, showing their markedly varied SAM architecture.
(c) Bar diagrams depicting the measurement of seven major SAM morphometric trait parameters (SWi, SMWi, SH, SH/SWi, SA, SAL, and SR) between two
accessions representing each desi (ICC 5590 and ICC 6306) and kabuli (ICC 12968 and ICC 16814) chickpea. Dotted lines represent variations among four desi
and kabuli accessions as well as between two accessions each belonging to desi (yellow) and kabuli (green). Significance at *P < 0.05 (pair-wise t-test) and
**P < 0.01 (pair-wise t-test).
(d) A hierarchical cluster display depicting the phenotypic correlation among SAM morphometric traits in the accessions, especially those representing desi
chickpea. The average Pearson’s correlation value estimated among the studied traits represented at the top with a colour scale; blue, dark blue and yellow
denote a low, medium and high levels of correlation, respectively. Digits indicated in the vertical and horizontal bars illustrate the range (minimum, optimum
and maximum) of the correlation coefficient measured among traits.
© 2019 The AuthorsThe Plant Journal © 2019 John Wiley & Sons Ltd, The Plant Journal, (2019), 98, 864–883
Transcriptional tags enhance chickpea productivity 867
To further assure the association potential of GWAS-
derived genomic SNP loci in governing PW, PH, SWi, SH,
and SA traits of chickpea, a gene-by-gene regional associa-
tion analysis was performed. For this process, high-resolu-
tion targeted resequencing at the most 150-kb genomic
regions (exhibiting significant LD decay) flanking six
GWAS-derived trait-associated SNPs was performed in 291
desi and kabuli accessions (association panel). The geno-
typing data for the SNPs from these target intervals of
associated genomic loci were integrated with the multien-
vironment (field/controlled condition) phenotyping data of
five studied plant and SAM architectural traits among
accessions belonging to an association panel. This proce-
dure delineated a shortest 0.14-Mb genomic interval (0.25–0.39 Mb with 16 genes) of significant LD resolution (mean
r2 = 0.86) covering either side of a GWAS-derived non-
synonymous coding SNP (CaKSNP3493(G/A) at 0.28 Mb) in
a CabHLH121 TF gene, which was strongly associated with
the PW, PH, SWi, SH, and SA traits (Figure 3b). This
finding infers the added-advantage of combining GWAS
and regional association analysis to scale-down a signifi-
cant trait-associated longer LD-block with many candidate
genes into a functionally relevant real causal gene, which
has been well demonstrated in multiple crop plants (Yano
et al., 2016).
Genetic mapping of major QTL ascertains the strong
association potential of CabHLH121 TF gene for major
plant architectural and SAM morphometric traits in
chickpea
Two accessions representing each desi (ICC 5590 and ICC
6306) and kabuli (ICC 12968 and ICC 16814) were selected
from the constituted association panel based on their con-
trasting PH and PW traits, as well as SAM morphometric
trait characteristics, for genetic mapping of major QTLs
governing plant and SAM architectural traits in chickpea
(Figure 1a,b). In desi chickpea, accession ICC 6306 with
broad PW and tall PH exhibited significantly higher trait
Figure 2. Phenotypic variation of major SAM morphometric and plant architectural traits in a constituted association panel.
(a) Frequency distribution of two major plant architectural traits, plant width (PW) and plant height (PH) variation, estimated among all 291, including 189 desi
and 102 kabuli, accessions.
(b) Boxplots depicting the phenotypic variation among three major SAM morphometric trait parameters, SWi, SH and SA, among all 291, including 189 desi and
102 kabuli, accessions. Box edges represent the upper and lower quantiles, with median values shown in the middle of the box.
(c) A hierarchical cluster display depicting the phenotypic correlation among plant architectural (PW and PH) and SAM morphometric (SWi, SH and SA) traits in
all 291, including 189 desi and 102 kabuli, accessions. The average Pearson’s correlation value measured among the studied traits is represented at the top with
a colour scale; blue, dark blue and yellow denote a low, medium and high levels of correlation, respectively. Digits indicated in the vertical and horizontal bars
illustrate the range (minimum, optimum and maximum) of correlation coefficient estimated among traits.
© 2019 The AuthorsThe Plant Journal © 2019 John Wiley & Sons Ltd, The Plant Journal, (2019), 98, 864–883
868 Laxmi Narnoliya et al.
values of major SAM morphometric trait parameters,
including SWi, SH, SMWi, SH/SWi, SAL, SA, and SR, com-
pared with those estimated in another narrow PW and
dwarf PH accession, ICC 5590 (Figure 1a–c). In kabuli chick-
pea, we could not observe any such significant variation in
SAM morphometric features other than SH, SH/SWi, and
SA (Figure 1c). The phenotypic trait diversity/correlation
observed among these four analysed desi and kabuli
accessions, including an association panel (291 acces-
sions), revealed a high significantly positive correspon-
dence between SAM morphometric and plant architectural
traits in desi chickpea. Therefore, two contrasting desi
accessions (ICC 5590 and ICC 6306) were selected to
develop a desi RIL mapping population (ICC 5590 9 ICC
6306), which was utilized further to construct a genetic link-
age map and QTL mapping in chickpea.
Primarily, a desi genetic linkage map (ICC 5590 9 ICC
6306) was constructed by integrating 9320 genome-wide
SNPs across eight chromosomes of chickpea. This gener-
ated a genetic map covering a total map length of
1085.98 cM with a map density (mean inter-marker dis-
tance) of 0.12 cM (Tables S4 and S5 and Figure S3).
Chromosomes 4 and 2 had the highest and lowest satu-
rated genetic maps with map densities of 0.08 and
0.17 cM, respectively. Therefore, a high-resolution genetic
linkage map can be efficiently utilized as a reference for
molecular mapping of major QTLs governing vital agro-
nomic traits in chickpea. The comprehensive multienvi-
ronment (field and controlled conditions) replicated
phenotyping of a RIL mapping population (ICC
5590 9 ICC 6306) exhibited a wider level of phenotypic
variation (6.62–13.83% CV) and high H2 (76–82%) for two
major plant architectural traits (PW and PH), as well as
three vital SAM morphometric traits (SWi, SH and SA),
among 278 mapping individuals and parental accessions
(Table S2 and Figure S4). Notably, one of the RIL map-
ping parental desi accessions, ICC 5590, exhibited a nar-
row PW (50.58 cm)/SWi (93 lm) and a low SA
(7476.81 lm2), with a dwarf PH (38.03 cm)/SH (86.78 lm).
Another RIL mapping parental desi accession, ICC 6306,
had a broad PW (68.07 cm)/SWi (119.99 lm) and high SA
(18 595.91 lm2), with a tall PH (66.34 cm)/SH (144.28 lm).
The normal frequency distribution and the bi-directional
transgressive segregation of the PW, PH, SWi, SH, and
SA traits were evident in a RIL mapping population (Fig-
ure S4). The detailed correlation analysis and subsequent
hierarchical cluster display depicted a positive correlation
of 81% among these plant and SAM architectural traits in
a RIL population (Figure S5).
For molecular mapping of major QTLs, the genotyping
data of 9320 SNPs genetically mapped on chromosomes
were correlated with the multienvironment replicated phe-
notyping information for PW, PH, SWi, SH and SA traits
among RIL mapping individuals and parental accessions.Table
1SNPssignifica
ntlyas
sociated
withmajorplantarch
itec
turalan
dSAM
morphometrictraits
inch
ickp
ea
SNPIDs
SNPs
SNPphys
ical
positions
(bp)
Kab
uli
chromoso
mes
Gen
eac
cession
IDs
Gen
eiden
tities
Structural
annotation
PCombined
PVE
(R2%)
Associated
traits
Cak
SNP13
40(C/G
)445
238
0Ca_
Kab
uli_
Chr04
––
Intergen
ic1.99
10�8
21PW,SWian
dSA
Cak
SNP14
42(A
/G)
1123
113
7Ca_
Kab
uli_
Chr04
Ca_
0435
5Majorintrinsicprotein
Non-syn
onym
ous(CDS)
2.59
10�8
19PW,PH,SWi,SH
andSA
Cak
SNP15
06(G
/C)
1378
772
0Ca_
Kab
uli_
Chr04
Ca_
0460
3Copper
amineoxidas
eNon-syn
onym
ous(CDS)
2.39
10�8
23PW,PH,SWi,SH
andSA
Cak
SNP15
12(G
/T)
1384
134
0Ca_
Kab
uli_
Chr04
Ca_
0460
8Unkn
ownex
pressed
protein
Non-syn
onym
ous(CDS)
2.09
10�7
15PH,SH
andSA
Cak
SNP29
69(G
/A)
5738
222
0Ca_
Kab
uli_
Chr06
Ca_
1370
5Zincfinger,RIN
G-typ
eNon-syn
onym
ous(CDS)
3.19
10�8
17PW,SWian
dSA
Cak
SNP34
93(G
/A)
28162
2Ca_
Kab
uli_
Chr08
Ca_
1190
9Bas
ichelix-loop-helix
(bHLH
)Non-syn
onym
ous(CDS)
1.09
10�9
34PW,PH,SWi,SH
andSA
SNPs,
single
nucleo
tidepolymorphisms;
SAM,sh
ootap
ical
meristem;PVE,phen
otypic
variationex
plained
;PW,plantwidth;PH,plantheight;SWi,SAM
width;SH,SAM
height;SA,SAM
area
.
© 2019 The AuthorsThe Plant Journal © 2019 John Wiley & Sons Ltd, The Plant Journal, (2019), 98, 864–883
Transcriptional tags enhance chickpea productivity 869
This exertion altogether identified four major genomic
regions harbouring 18 major QTLs (6.0–16.8 logarithm of
the odds, LOD) governing PW, PH, SWi, SH, and SA traits
that were mapped on three chickpea chromosomes (Ca04,
Ca06 and Ca08) (Table S6, Figure S3). The PVE determined
by individual QTLs for five studied architectural traits var-
ied from 10.8 to 41.8% R2. The PVE estimated for all 18
QTLs in combination was 44.7% R2. The detected 18 major
QTLs demonstrated additive effects for desirable PW, PH,
SWi, SH, and SA traits, reflecting the effective contribution
of favourable alleles derived from both RIL mapping paren-
tal desi accessions ICC 5590 and ICC 6306 on these loci to
modulate the plant and SAM architectural traits (Table S6).
The integration of QTL mapping with GWAS collec-
tively demonstrated that non-synonymous coding SNPs
in the four candidate genes linked to 18 major QTLs were
significantly associated with PW, PH, SWi, SH, and SA
traits (Tables 1 and S6 and Figures 3a,b and S3). Notably,
Figure 3. CabHLH121 strongly associated with five major plant and SAM architectural traits expresses specifically in the SAM tissue.
(a) GWAS-derived Manhattan plot showing significant P-values associated with five major plant architectural (PW and PH) and SAM morphometric (SWi, SH,
and SA) traits, obtained using the genome-wide SNPs mined from 291 desi and kabuli germplasm accessions of chickpea. The x-axis represents the relative den-
sity of SNPs that physically mapped on eight chromosomes and unanchored scaffolds of the kabuli chickpea genome. The y-axis indicates the �log10 (P)-value
for significant association of SNPs with the studied plant architectural and SAM morphometric traits. Six SNPs exhibiting significant association with traits at
cut-off of P ≤ 1 9 10�8 are marked with dotted lines.
(b) Regional association analysis-led high-resolution LD heat map covering a 0.14-Mb (0.25–0.39 Mb) genomic interval (highlighted with red dotted lines) surround-
ing a GWAS-derived non-synonymous (arginine: CGT to histidine: CAT) coding SNP (CaKSNP3493 (G/A) at 0.28 Mb) on chromosome 8 exhibited a strong association
with plant architectural (PW and PH) and SAMmorphometric (SWi, SH and SA) parameters. The arrow specifies the genomic position (bp) of the trait-associated SNP
on chromosome 8. In a LD heat map, r2 indicates the frequency correlation among a pair of alleles across a pair of SNP loci.
(c) MultiExperiment Viewer (MeV) illustrating the differential expression profiles of four significant plant and SAM architectural trait-associated genes in diverse tis-
sues, including germinating seedling (GS), young leaf (YL), and shoot apical meristem (SAM) as well as eight flower development stages (FB1, FB2, FB3, FB4, FL1,
FL2, FL3, and FL4) of desi chickpea accession ICC 4958. The average log signal expression values of genes in various tissues/stages are mentioned at the top with a
colour scale, in which green, black and red denote low, medium and high levels of expression, respectively. Tissues/development stages of ICC 4958 and accession
IDs of genes used for expression profiling are mentioned at the top and right-side of the expression map, respectively. FB: flower bud and FL: mature flower.
(d) Differential expression profiling of a strong plant and SAM architectural trait-associated CabHLH121 TF gene in the vegetative leaf, shoot and SAM tissues of two
parental desi accessions (ICC 5590 and ICC 6306) of a RIL mapping population (ICC 5590 9 ICC 6306), as well as in two kabuli accessions (ICC 12968 and ICC 16814),
contrasting with the plant architectural and SAM morphometric traits using the quantitative RT-PCR assay. Each bar denotes the mean (� standard error) of three
independent biological replicates with two technical replicates for each sample used in the RT-PCR assay. *Significant difference in gene expression estimated by the
LSD-ANOVA significance test at P < 0.05.
© 2019 The AuthorsThe Plant Journal © 2019 John Wiley & Sons Ltd, The Plant Journal, (2019), 98, 864–883
870 Laxmi Narnoliya et al.
one non-synonymous SNP in a CabHLH121 TF gene
mapped on the major genomic region (1.360 cM corre-
sponding to 281 622 bp) of CaqPW/PH/SWi/SH/SA8.1
QTLs exhibited strong association potential for PW, PH,
SWi, SH, and SA traits based on high-resolution GWAS
(10�9 P with 34% PVE) and QTL mapping (11.8–16.8 LOD
with 43.6% PVE) (Table 1, Table S6). At this major QTL
region, an additive interaction was observed for QTLs/al-
leles governing desirable traits derived from both map-
ping parental desi accessions ICC 5590 and ICC 6306. In
summary, informative non-synonymous SNP alleles of
CabHLH121 TF gene tightly linked to the major QTLs
(CaqPW/PH/SWi/SH/SA8.1) governing PW, PH, SWi, SH,
and SA traits were successfully validated by high-resolu-
tion GWAS, regional association analysis and QTL map-
ping in chickpea. Next, these functionally relevant
molecular tags were selected as potential candidates for
efficient dissection of plant architectural and SAM mor-
phometric traits in chickpea.
Transcript profiling assures the regulatory function of a
strong trait-associated CabHLH121 TF gene for major plant
architectural and SAM morphometric traits in chickpea
Four significant PW, PH, SWi, SH, and SA trait-associated
genes, including a strong trait-associated CabHLH121 TF
validated by GWAS, regional association analysis and QTL
mapping, were assessed for their differential regulatory
function in controlling major plant architectural and SAM
morphometric traits of chickpea. For this procedure, in sil-
ico global transcriptome profiling of said genes was per-
formed in germinating seedlings, young leaves and SAM
as well as four development stages of each immature
flower bud and mature flower tissues of desi accession ICC
4958. All four trait-associated genes showed SAM-specific
expression and were differentially upregulated (≥5-fold at
P ≤ 0.01), especially in the SAM tissue, compared with ger-
minating seedlings, young leaves, flower buds, and mature
flowers of ICC 4958 (Figure 3c). Of these, a non-synon-
ymous SNP-carrying CabHLH121 TF gene associated
strongly with PW, PH, SWi, SH, and SA traits and exhibit-
ing pronounced SAM-specific expression was upregulated
more than 12-fold in the SAM compared with the other
analysed tissues of ICC 4958 (Figure 3c). Therefore,
CabHLH121 TF was considered a promising candidate gene
regulating two major plant architectural traits (PW and PH),
possibly through SAM proliferation and differentiation in
chickpea.
To experimentally validate this strong trait-associated
CabHLH121 TF gene, its differential expression profiling
was performed in the shoots, young leaves and SAM of
two RIL mapping parental desi accessions (ICC 5590 and
ICC 6306), as well as in two additional kabuli accessions
(ICC 12968 and ICC 16814), contrasting with the PW, PH,
SWi, SH, and SA traits assessed by quantitative RT-PCR
assay. The CabHLH121 TF gene with a non-synonymous
SNP showed SAM-specific expression and was signifi-
cantly upregulated (15.2-fold) in the SAM compared with
that in the respective shoots and young leaves of four said
desi and kabuli accessions (Figure 3d). A pronounced
upregulation (5.2-fold) of CabHLH121 in the SAM of a
broad PW/SWi and tall PH/SH desi mapping parental
accession ICC 6306 with enhanced SAM morphometric
characteristics was apparent compared with that of its
counterpart narrow PW/SWi and dwarf PH/SH desi acces-
sion ICC 5590 (Figure 3d). However, only a 1.7-fold upregu-
lation of CabHLH121 was observed in broad PW/SWi and
tall PH/SH kabuli accession ICC 16814 in comparison with
its counterpart kabuli accession ICC 12968 with a narrow
PW/SWi and dwarf PH/SH. When we compared the desi
and kabuli accessions collectively, significant upregulated
expression of CabHLH121 was observed, especially in the
SAM of a broad PW/SWi and tall PH/SH desi accession
(ICC 6306) compared with a broad PW/SWi and tall PH/SH
kabuli accession (ICC 16814) (Figure 3d). Overall, this find-
ing infers that the degree of CabHLH121 gene expression
was enhanced significantly with the subsequent increase
in PW, PH, SWi, SH, and SA, especially in the accessions of
desi chickpea, suggesting a strong positive correlation
between the gene expression level and varying plant/SAM
architectural trait parameters.
Natural haplotypes of a CabHLH121 TF gene govern major
plant architectural and SAM morphometric traits in
chickpea
A strong PW, PH, SWi, SH, and SA trait-associated
CabHLH121 TF gene showing SAM-specific upregulated
expression, validated by GWAS, regional association anal-
ysis, QTL mapping and expression profiling, was
sequenced among 291 desi and kabuli accessions (associa-
tion panel) for molecular haplotyping. This task identified
35 SNPs, including 16 upstream regulatory SNPs as well as
five missense and three non-sense non-synonymous and
four synonymous coding SNPs, which altogether consti-
tuted two major haplotypes (HAP A and HAP B) (Figure 4a,
b). The gene haplotype-based association analysis was
performed by integrating the genotyping information for
the SNP haplotypes in CabHLH121 gene with multienviron-
ment PW, PH, SWi, SH and SA trait phenotyping data
among accessions of an association panel. This procedure
revealed a strong association of haplotype HAP A (10�11 P
with 36.7% PVE) with narrow PW (45–54 cm by 18 acces-
sions)/dwarf PH (35–45 cm by 18 accessions)/narrow SWi
(88–97 lm by 24 accessions)/dwarf SH (73–100 lm by 24
accessions)/less SA (6478–13 672 lm2 by 26 accessions)
traits, referred to here as DNDL (desi narrow PW/SWi,
dwarf PH/SH, less SA). By contrast, haplotype HAP B (10�12 P
with 38.5% PVE) exhibited a strong association with broad
PW (55–69 cm by 56 accessions)/tall PH (46–66 cm by 56
© 2019 The AuthorsThe Plant Journal © 2019 John Wiley & Sons Ltd, The Plant Journal, (2019), 98, 864–883
Transcriptional tags enhance chickpea productivity 871
accessions)/broad SWi (96–120 lm by 50 accessions)/tall
SH (107–156 lm by 58 accessions)/high SA (14 270–18 569 lm2 by 48 accessions) traits, hereafter referred to
as DBTH (desi broad PW/SWi, tall PH/SH, high SA) (Fig-
ure 4c). In summary, the integrated genomic strategy
involving molecular haplotyping and haplotype-specific
association mapping delineated functionally relevant novel
alleles and natural haplotypes (DNDL and DBTH) of
CabHLH121 TF gene regulating PW and PH traits by modu-
lating the major SAM morphometric traits (SWi, SH and
SA) in chickpea.
Fine-mapping and map-based cloning ascertains the role
of a CabHLH121 TF gene harbouring a major QTL in the
regulation of vital plant architectural and SAM
morphometric traits of chickpea
One strong 1.356 cM major QTL (CaqPW/PH/SWi/SH/
SA8.1) interval (CaSNP8902(G/T) (1.058 cM) to CaSNP8909
(A/C) (2.414 cM)) harbouring a CabHLH121 TF, mapped on
chromosome 8 of a desi RIL genetic linkage map (ICC
5590 9 ICC 6306) was selected for fine-mapping to assure
the potential of this gene in governing PW and PH traits of
chickpea. For this process, a 1.356 cM genomic region cor-
responding to the indicated QTL interval (280.862 kb) from
the parental accessions (ICC 5590 and ICC 6306) was intro-
gressed and further backcrossed as recurrent and donor
parents four times between each other to produce two
BC4F4 NILs, DNDL-NILCaqPW/PH/SWi/SH/SA8.1 and DBTH-NIL-
CaqPW/PH/SWi/SH/SA8.1 with approximately 81% recovery
of the recurrent parental genome (Figure S6). Using 168
mapping individuals of a F2 population (DNDL-NILCaqPW/
PH/SWi/SH/SA8.1 9 DBTH-NILCaqPW/PH/SWi/SH/SA8.1), 17
recombinants were detected between CaSNP8903(A/C)
(1.206 cM) and CaSNP8908(A/T) (1.721 cM) SNPs at the
0.515 cM QTL interval (Figure 5a). This delineated genomic
interval underlying CaqPW/PH/SWi/SH/SA8.1 QTLs
mapped on chromosome 8 was defined into the 106.641-kb
region by integrating its genetic linkage map information
with that of the physical map of the kabuli chickpea
genome (Figure 5a).
The high-coverage (~1179) multiplex amplicon rese-
quencing of a 106.641-kb QTL genomic region in the paren-
tal accessions and two selected DNDL and DBTH
homozygous individuals of a mapping population (DNDL-
NILCaqPW/PH/SWi/SH/SA8.1 9 DBTH-NILCaqPW/PH/SWi/SH/SA8.1)
together detected 605 SNPs with 18 protein-coding genes
(Table S7). The QTL region-specific association analysis at
a CaqPW/PH/SWi/SH/SA8.1 QTL genomic interval was per-
formed by integrating the genotyping information of 605
SNPs with the multienvironments (field/controlled condi-
tion) phenotyping data of the PW, PH, SWi, SH, and SA
traits among 291 desi and kabuli accessions (association
panel). This approach detected a strong plant and SAM
architectural trait-associated CabHLH121 TF gene mapped
in the 6.189-kb QTL genomic interval between CaS214(A/G)
and CaS256(G/A) SNPs. High-resolution QTL mapping in
the RILs as well as QTL-introgressed NILs and their integra-
tion with QTL region-specific association analysis overall
scaled-down the 106.641-kb CaqPW/PH/SWi/SH/SA8.1 QTL
interval into a 6.189-kb genomic region (Figure 5b). By the
comprehensive phenotyping of 17 recombinants following
progeny testing, a 106.641-kb QTL genomic interval was
resolved into the 6.189-kb region between CaS214(A/G)
and CaS256(G/A) SNPs in seven selected most-promising
recombinants of NILs (Figure 5b). The structural and func-
tional annotation of this 6.189-kb QTL genomic interval
with the kabuli genome identified one protein-coding gene,
CabHLH121 TF with accession ID Ca_11909 (Figure 5b). A
non-synonymous SNP (G/A) that was tightly linked to a
CabHLH121 TF and exhibiting zero recombination with this
target locus in seven selected recombinants of NILs was
considered the most-promising gene governing PW, PH,
SWi, SH, and SA traits in chickpea. The same gene delin-
eated through integrating high-resolution GWAS, regional
association analysis and QTL mapping with molecular hap-
lotyping and expression profiling was further validated by
fine-mapping and map-based cloning in the RILs and NILs.
Therefore, the CabHLH121 TF gene harbouring a CaqPW/
PH/SWi/SH/SA8.1 major QTL indeed surfaced as the most
potential candidate regulating PW and PH primarily by
modulating major SAM morphometric traits (SWi, SH, and
SA) in chickpea.
Development of NILs introgressed with superior
CabHLH121 gene haplotypes to understand their
regulatory function in influencing major plant architectural
and SAM morphometric traits of chickpea
For haplotype-assisted foreground selection, the SNPs
flanking (CaS214(A/G) (277.582 kb) and CaS256(G/A)
(283.771 kb)) and linked (CaKSNP3493(G/A) (281.622 kb))
to the 6.189-kb major CaqPW/PH/SWi/SH/SA8.1 QTLs and
the strong trait-associated CabHLH121 TF gene haplo-
types (DNDL and DBTH) were used (Figures 4d and S6).
Subsequently, 10 positive lines screened by foreground
selection were evaluated for background selection with
the genome-wide 1536 parental (ICC 5590 and ICC 6306)
polymorphic SNPs uniformly mapped on eight chromo-
somes. These procedures enhanced the recovery of the
parental recurrent genome up to 97.2–98.0% in the con-
stituted DNDL and DBTH gene (CabHLH121 TF) haplo-
type-introgressed NILs, DNDL-NILCabHLH121(HAPA) and
DBTH-NILCabHLH121(HAPB). This result infers the efficacy of
gene haplotype-assisted foreground/background selection
compared with the most commonly used marker-assisted
selection in the recovery of the parental recurrent gen-
ome as well as the precise selection of the most-promis-
ing recombinants for further genetic enhancement
studies in chickpea.
© 2019 The AuthorsThe Plant Journal © 2019 John Wiley & Sons Ltd, The Plant Journal, (2019), 98, 864–883
872 Laxmi Narnoliya et al.
The CabHLH121 TF gene haplotype-specific differential
expression profiling depicted a higher expression (≥6-fold)and enhanced accumulation of transcript of the DNDL gene
haplotype in the SAM of corresponding haplotype-intro-
gressed NILs (DNDL-NILCabHLH121(HAPA)) (Figure 6a). How-
ever, their counterpart DBTH-introgressed NILs (DBTH-
NILCabHLH121(HAPB)) had lower expression and a reduced
transcript accumulation of the DBTH gene haplotype.
The efficiency of CabHLH121 gene haplotypes (DNDL
and DBTH) in modulating plant/SAM architectural traits
was assured based on the comprehensive phenotypic
evaluation and characterization of corresponding haplo-
type-introgressed NILs (DNDL-NILCabHLH121(HAPA) and
DBTH-NILCabHLH121(HAPB)). For this process, SAM mor-
phometric parameters, including SWi, SH, SA, SAL, SMWi,
and SR, were measured in these NILs along with RIL
mapping parental desi accessions (ICC 5590 and ICC 6306)
through histology. The results demonstrated high signifi-
cant trait values of all these SAM morphometric features in
the DBTH-NILCabHLH121(HAPB) line compared with those esti-
mated in the DNDL-NILCabHLH121(HAPA) line (Figure 7a,b).
The field phenotypic evaluation and scanning electron
microscopy (SEM) of these NILs along with mapping par-
ental accessions revealed that the DBTH-NILCabHLH121(HAPB)
line with a broad PW and tall PH had a 1.7 to 2-fold higher
SA compared with the DNDL-NILCabHLH121(HAPA) line with a
narrow PW and dwarf PH (Figure 8a,b). A comprehensive
correlation analysis and subsequent hierarchical cluster
display demonstrated a significant positive correlation (88–94%) among SAM morphometric and plant architectural
traits (PW, PH, SWi, SH, SA, SAL, and SR) in the developed
NILs (Figure 7c).
(b)
(a)
(d)(c)
Figure 4. Molecular haplotyping of CabHLH121 TF and development of gene haplotype-introgressed NILs.
(a, b) Genotyping of 35 SNPs mined from coding as well as non-coding introns and 3 kb of each URR and DRR sequence component of a CabHLH121 TF gene
among 291 desi and kabuli chickpea accessions (association panel) constituting two major haplotypes, HAP A (DNDL) and HAP B (DBTH). CDS: coding DNA
sequence, URR/DRR: upstream/downstream regulatory region.
(c) Boxplots illustrating the significant phenotypic variation of five major plant architectural (PW and PH) and SAM morphometric (SWi, SH and SA) traits among
291 desi and kabuli chickpea accessions. Box edges denote the lower and upper quantiles, with median values shown in the middle of the box. Digits within the
square parentheses indicate the number of desi and kabuli chickpea accessions represented by the individual haplotypes HAP A (DNDL) and HAP B (DBTH).
(d) Development of CabHLH121 gene haplotype-introgressed NILs, DNDL-NILCabHLH121(HAPA) and DBTH-NILCabHLH(HAPB) by marker (haplotype)-assisted foreground
and background selection using two mapping parental accessions (ICC 5590 and ICC 6306), contrasting with the PW, PH, SWi, SH, and SA traits. These developed
NILs were compared with a high-yielding Indian desi variety, ICCV 93954, to evaluate their overall agronomic performance, especially targeting five studied plant
architectural and SAM morphometric traits in the field and controlled environment. DNDL: Desi narrow plant/SAM width, dwarf plant/SH, less SAM area. DBTH:
Desi broad plant/SAM width, tall plant/SAM height, high SAM area.
© 2019 The AuthorsThe Plant Journal © 2019 John Wiley & Sons Ltd, The Plant Journal, (2019), 98, 864–883
Transcriptional tags enhance chickpea productivity 873
The promoter of CabHLH121 interacts with CaWUS TF, a
master regulator of stem cell niche
The molecular basis of the differential expression of a
strong trait-associated SAM-specific CabHLH121 TF and its
constituted DNDL and DBTH gene haplotypes in influenc-
ing major plant architectural (PW and PH) and SAM mor-
phometric (SWi, SH, and SA) traits was determined. To
achieve this goal, the 3-kb upstream regulatory region
(URR) of a CabHLH121 gene was scanned primarily for the
presence of known cis-regulatory elements/TF binding
sites. This procedure detected an abundance of AT-hook
and GATA binding sequences comprising 14–21% of the
total identified TF binding sites in the URR of a CabHLH121
gene in chickpea (Figure S7). Interestingly, 5% of these
mined known cis-regulatory element regions contained
WOX-homeodomain TF binding sites in the 3-kb URR of a
CabHLH121 gene. Comprehensive in silico analysis further
demonstrated that a homeodomain TF gene (Ca_01974)
encoding WUSCHEL (CaWUS) in chickpea, orthologous to
the Arabidopsis AtWUS gene (At2g17950), potentially
bound to these WOX-homeodomain TF binding sites in the
URR (promoter) of CabHLH121. Therefore, this identified
CaWUS, a master regulator of stem cell maintenance, pro-
liferation and differentiation of SAM in multiple crop
plants, may directly target and essentially regulate the
expression of a CabHLH121 TF gene in chickpea (Leiboff
et al., 2015; Thompson et al., 2015; Lee et al., 2016).
To ascertain the binding of CaWUS to the TF binding
sites of the URR of CabHLH121, both chromatin immuno-
precipitation (ChIP)-seq and ChIP-qPCR assays were
Figure 5. Fine-mapping andmap-based cloning narrowed-down a major QTL into CabHLH121 TF gene controlling major plant and SAM architectural traits in chickpea.
(a) Map-based cloning of a major CaqPW/PH/SWi/SH/SA8.1 QTL by fine-mapping in a NIL mapping population (DNDL-NILCaqPW/PH/SWi/SH/SA8.1 9 DBTH-NILCaqPW/PH/
SWi/SH/SA8.1) and subsequent QTL region-specific association analysis in 291 desi and kabuli chickpea accessions contrasting with plant architectural and SAM mor-
phometric traits (PW, PH, SWi, SH, and SA). The identity of the markers mapped on the linkage groups (LGs)/chromosomes and their genetic (cM)/physical (bp) posi-
tions are indicated above and below the chromosomes, respectively. The SNPs flanking and tightly linked to the QTLs governing the PW, PH, SWi, SH, and SA traits
are indicated in blue and red colour, respectively.
(b) Progeny testing-based individual phenotyping of seven selected recombinants as well as contrasting NILs (DNDL-NILCaqPW/PH/SWi/SH/SA8.1 and DBTH-NILCaqPW/PH/
SWi/SH/SA8.1) and mapping parental accessions (ICC 5590 and ICC 6306) to deduce the genotype of a 6.189-kb delineated major CaqPW/PH/SWi/SH/SA8.1 QTL in chick-
pea. A non-synonymous SNP, CakSNP3493 (G/A), tightly linked to the CabHLH121 TF gene and exhibiting zero recombination in seven selected recombinants was
associated strongly with PW, PH, SWi, SH, and SA traits in chickpea. The genomic constitution of DNDL (Desi narrow plant/SAM width, dwarf plant/SAM height, less
SAM area) and DBTH (Desi broad plant/SAMwidth, tall plant/SAM height, high SAM area) NILs/parental accessions are represented by ‘A’ and ‘B’, respectively.
© 2019 The AuthorsThe Plant Journal © 2019 John Wiley & Sons Ltd, The Plant Journal, (2019), 98, 864–883
874 Laxmi Narnoliya et al.
performed (Figure 6c–e). ChIP was performed in the DNDL-
NILCabHLH121(HAPA) and DBTH-NILCabHLH121(HAPB) lines by
immunoprecipitation of the chromatin complex with a
WUS-specific antibody and IgG (immunoglobin G) as a
control. The sequencing of immunoprecipitated DNA and
subsequent analysis of ChIP-seq data exhibited 17.8X gen-
ome coverage of the kabuli chickpea genome. Mapping of
the sequence coverage at each physical position (bp) on
the chickpea genome generated peaks with an abundant
genomic distribution in the 3-kb regulatory region
upstream of the translation initiation codon (ATG) of the
CabHLH121 gene. Genes in the URRs exhibiting similar
peaks with a height <90 in the control were removed, and
the remaining 47 genes with a peak height ≥90 were
screened. The ChIP-seq outcome was further correlated
with the known cis-regulatory element regions/TF binding
sites predicted in the URR of a CabHLH121 and its consti-
tuted DNDL and DBTH gene haplotypes (Figure 6c). This
procedure identified CaWUS (Ca_01974) as a candidate TF
with a maximum peak height of 267 that bound to the 594-
686-bp region containing two predicted successive WOX-
homeodomain TF binding sites in the URR of CabHLH121
and both of its DNDL and DBTH gene haplotypes (Fig-
ure 6d). The structural annotation of these concurrent
WUS TF binding sites in a CabHLH121 gene revealed their
correspondence to the 280262–280269 and 280347–280354-bp pseudomolecule sequence physical positions on chro-
mosome 8. The ChIP-qPCR assay using the chromatin
Figure 6. CabHLH121 is transcriptionally regulated by CaWUS, a homeobox TF.
(a, b) Differential expression profiling of the delineated CabHLH121 and CaWUS genes in the vegetative leaf, shoot and SAM tissues of two parental desi acces-
sions (ICC 5590 and ICC 6306) of a RIL mapping population and developed NILs (DNDL-NIL and DBTH-NIL) using the quantitative RT-PCR assay. Each bar denotes
the mean (� standard error) of three independent biological replicates with two technical replicates for each sample used in the RT-PCR assay. *Significant differ-ences in gene expression measured by an LSD-ANOVA significance test at P < 0.05.
(c) Gene structural annotation illustrating the accurate position (bp) of binding of a CaWUS to the WOX-homeodomain TF binding sites in the 3-kb upstream regu-
latory (promoter) region of a CabHLH121 TF gene, determined by DNA-protein interaction assays (ChIP-seq and ChIP-qPCR). CDS: coding DNA sequence and FD:
functional domain.
(d, e) ChIP-seq was performed using the DNDL-NIL and DBTH-NIL lines by immunoprecipitation of the chromatin complex with an anti-WUS antibody and using
immunoglobin G (IgG) as a negative control, followed by sequencing of the immunoprecipitated DNA.
(d) ChIP-seq derived peak with a highest peak height of 267 revealed the binding of CaWUS to the concurrent WOX-homeodomain TF binding sites in the upstream
regulatory (promoter) regions (280 262–280 269 bp and 280 347–280 354 bp) of the CabHLH121 TF gene.
(e) The ChIP-qPCR assay confirmed the binding of CaWUS to the successive WUS TF binding sites in the aforesaid upstream regulatory region of the CabHLH121
gene. ChIP-qPCR was performed to amplify the immunoprecipitated DNA with forward/reverse primers (indicated by arrows) targeting the WUS TF binding sites
in CabHLH121. The ChIP-qPCR results are presented as fold changes by dividing the signals from ChIP with anti-WUS antibody by the IgG control. Bars denote the
mean (� standard error). For a simpler representation, the CabHLH121 gene haplotype-introgressed NILs, DNDL-NILCabHLH121(HAPA) and DBTH-NILCabHLH121(HAPB)
are coded with DNDL-NIL and DBTH-NIL, respectively.
© 2019 The AuthorsThe Plant Journal © 2019 John Wiley & Sons Ltd, The Plant Journal, (2019), 98, 864–883
Transcriptional tags enhance chickpea productivity 875
immunoprecipitated DNA from the DNDL-NILCabHLH121
(HAPA) and DBTH-NILCabHLH121(HAPB) lines was performed
with gene-specific primers. This process revealed the bind-
ing of a CaWUS to the WOX-homeodomain TF binding
sites in the URR of a CabHLH121 gene/haplotype (DNDL
and DBTH) in controlling their transcriptional regulation for
differential SAM morphometric characteristics and plant
architectural traits of chickpea (Figure 6e). Furthermore,
differential expression profiling of a CaWUS revealed
lower expression (≥7-fold) and a reduced accumulation
of its transcript in the SAM of the DNDL-NILCabHLH121(HAPA)
line (Figure 6b). However, their counterpart
DBTH-NILCabHLH121(HAPB) line showed higher expression
and enhanced accumulation of the CaWUS transcript.
Phenotypic evaluation and characterization of CabHLH121
gene haplotypes-introgressed NILs for vital agronomic
traits
The effects of the CabHLH121 gene haplotypes (DNDL and
DBTH) on major plant architectural and SAM
morphometric traits were compared between NILs (DNDL-
NILCabHLH121(HAPA) and DBTH-NILCabHLH121(HAPB)), along with
RIL mapping parental desi accessions (ICC 5590 and ICC
6306) and a high-yielding Indian desi chickpea variety
(ICCV 93954), based on their multienvironment replicated
phenotypic evaluation in the field. Compared with the
other lines, the DNDL-NILCabHLH121(HAPA) line exhibited an
optimal PW (47.6 cm) and desirable semidwarf PH
Figure 7. Phenotypic variation of major SAM morphometric and plant architectural traits in the developed NILs.
(a) SAM architecture of the DNDL-NIL and DBTH-NIL lines examined based on major morphometric trait parameters (SWi, SH, SMWi, SR, SAL and SA) using a
histological assay, indicating their markedly altered SAM architecture.
(b) Bar diagrams illustrating the measurement of plant architectural and SAM morphometric traits between DNDL-NIL and DBTH-NIL lines. Dotted lines repre-
sent significant variation among DNDL-NIL and DBTH-NIL lines. Significance at a *P < (pair-wise t-test) and **P < (pair-wise t-test). The CabHLH121 gene haplo-
type-introgressed NILs, DNDL-NILCabHLH121(HAPA) and DBTH-NILCabHLH121(HAPB) are indicated by DNDL-NIL and DBTH-NIL, respectively.
(c) A hierarchical cluster display depicting the phenotypic correlation among major plant architectural (PW and PH) and SAM morphometric (SWi, SH, SMWi,
SR, SAL and SA) traits in the NILs. The mean Pearson’s correlation value estimated among the traits is represented at the top with a colour scale; blue, dark blue
and yellow denote low, medium and high levels of correlation, respectively. Digits indicated in the vertical and horizontal bars illustrate the range (minimum,
optimum and maximum) of correlation coefficient among the studied traits.
© 2019 The AuthorsThe Plant Journal © 2019 John Wiley & Sons Ltd, The Plant Journal, (2019), 98, 864–883
876 Laxmi Narnoliya et al.
(40.5 cm) without compromising its multiple other desir-
able plant architectures, agromorphological, and seed yield
trait attributes. These useful agronomic trait characteristics
included a semierect plant growth habit, yellow brown
seed colour, and early flowering (48 days) and maturity
(101 days) time, as well as an increased branch and pod/
seed number and enhanced seed weight, yield per plant
and yield per hectare (productivity), among others (Fig-
ure 4d, Table S8). The developed superior NILs (DNDL-NIL-
CabHLH121(HAPA)) by contrast with the other analysed
NILs and accessions together showed 26, 31.2, and 31.7%
decreases in flowering/maturity time, PH and PW, respec-
tively, with a 13.3% increase in yield/productivity. This find-
ing implicates the greater efficacy of the superior DNDL
haplotype delineated from a CabHLH121 TF gene in impart-
ing optimal PW and PH without undesirable pleiotropic
effects on other agromorphological and yield component
traits. Henceforth, these functionally relevant molecular
tags will be useful to develop high-yielding new plant type
cultivars that are restructured with a desirable PH and PW
through the modulation of vital plant architectural and
SAM morphometric traits in chickpea.
DISCUSSION
Efficient genetic manipulation of various major plant archi-
tectural traits is vital to achieve enhanced yield and
productivity in crop plants (Pickersgill, 2007; Jin et al.,
2008; Wang and Li, 2008). Plant architectural traits, includ-
ing PH and width, are modulated by differentiation, prolif-
eration and maintenance of the meristematic stem cell
population, which can be primarily assessed by quantita-
tive measurements of diverse SAM morphometric trait
parameters, such as SAM volume/size (height, width, and
area), in crops (Thompson et al., 2014, 2015; Leiboff et al.,
2015; Cai et al., 2016; Saxena et al., 2017). From this per-
spective, this study evaluated the multienvironment (field/
controlled condition) phenotypic variation in two vital plant
architectural traits (PW and PH) and correlated the findings
with seven major SAM morphometric trait parameters (SH,
SWi, SMWi, SH/SWi, SA, SAL, and SR) in the constituted
association panel (291 desi and kabuli germplasm acces-
sions) and RIL/NIL mapping individuals of chickpea
through histological and SEM assays. A strong positive
correlation (91–93%) of PW and PH with the SAM morpho-
metric trait characteristics was evident among accessions/
individuals, especially for desi chickpea. However, this
observed definite correlation between SAM morphometric
and plant architectural traits was not valid among acces-
sions of kabuli chickpea, potentially due to distinct differ-
ences in the genetic backgrounds as well as striking
variation in the agromorphological characteristics between
desi and kabuli chickpea (Varshney et al., 2013a,b;
Figure 8. Correlation of PW with SAM morphometric parameters regulated by interplay between CabHLH121 and CaWUS.
(a) SAM architecture of two desi accessions (ICC 5590 and ICC 6306) of an RIL mapping population and two NIL lines (DNDL-NIL and DBTH-NIL) examined based
on a major SAM morphometric parameter using SEM, revealing a marked variation in the average SAM area (lm2) among these accessions/NILs. Comparative
overview model depicting the role of a CaWUS in the differential transcriptional regulation and repression of the CabHLH121 TF gene, thereby modulating the
average SAM area in the meristematic stem cell population of SAM representing the DNDL-NIL and DBTH-NIL lines. CaWUS and CabHLH121 are highlighted
with green and red colour based on their downregulated and upregulated expression, respectively, in the DNDL-NIL and DBTH-NIL lines.
(b) Replicated field phenotypic evaluation and characterization of DNDL-NIL and DBTH-NIL lines exhibiting striking differences in plant width (PW), as reflected
by their significant variation in the average area (lm2) under the curves. The CabHLH121 gene haplotype-introgressed NILs, DNDL-NILCabHLH121(HAPA) and DBTH-
NILCabHLH121(HAPB) are marked by DNDL-NIL and DBTH-NIL, respectively.
© 2019 The AuthorsThe Plant Journal © 2019 John Wiley & Sons Ltd, The Plant Journal, (2019), 98, 864–883
Transcriptional tags enhance chickpea productivity 877
Purushothaman et al., 2014; Penmetsa et al., 2016). The
detailed phenotypic evaluation and characterization of
large-scale natural and mapping populations collectively
inferred that major SAM morphometric trait parameters,
including SWi, SH, and SA, served as prime indicators to
define and understand the genetic basis of the broad phe-
notypic variation observed for PW and PH traits, particu-
larly among accessions of desi chickpea, and therefore can
be targeted in genomics-assisted breeding for the genetic
enhancement of chickpea. This finding further ascertains
the direct involvement of SAM morphometric features in
controlling the two major plant architectural traits (PW and
PH), possibly by influencing the differentiation and prolifer-
ation of the meristematic stem cell population in the SAM
of at least desi chickpea.
In this context, a comprehensive analysis of these plant
architectural (PW and PH) and SAM morphometric traits
was performed by integrating GWAS, regional association
analysis, QTL/fine-mapping, map-based cloning, and
molecular haplotyping with gene haplotype-specific associ-
ation and transcript profiling, as well as the protein�DNA
interaction assay, in the constituted association panel and
mapping populations of chickpea. This procedure rapidly
delineated natural allelic variants and haplotypes (DNDL
and DBTH) of a CabHLH121 TF gene within the major
CaqPW/PH/SWi/SH/SA8.1 QTL that regulates PW, PH, and
SAM morphometric traits primarily through the prolifera-
tion and differentiation of the SAM meristematic stem cell
population in desi chickpea. The identification of potential
molecular tags of a single CabHLH121 TF gene in imparting
differential SAM morphometric characteristics and plant
architectural traits owes to their strong phenotypic correla-
tion in natural and mapping populations of chickpea.
The CabHLH121 TF and its constituted gene haplotypes
(DNDL and DBTH) is transcriptionally regulated by a mas-
ter regulator, CaWUS, as evidenced by expression profiling
and protein�DNA interaction assays (ChIP-seq and ChIP-
qPCR) in the corresponding haplotype-introgressed NILs.
The WUS and some other WOX family TFs are key players
in maintaining the meristematic stem cell population in the
model plant Arabidopsis as well as many other crop plants
(Laux et al., 1996; Leibfried et al., 2005; Leiboff et al., 2015;
Thompson et al., 2015; Lee et al., 2016). WUS is known to
bind, coordinate and repress the transcriptional activity of
a bHLH TF gene, HECATE1 (HEC1), and thereby essentially
promote meristematic stem cell proliferation and SAM size
expansion in crop plants (Buechel et al., 2010; Schuster
et al., 2014; Sparks and Benfey, 2014). In this study, the
delineated bHLH TF CabHLH121 was found to be one of
the major TFs regulating both PH and PW traits simultane-
ously in chickpea. CabHLH121 acts downstream of a
CaWUS in the signalling cascade and is responsible for the
transcriptional regulation of major genes affecting SAM
morphometric traits in chickpea. An inverse relationship
between the expression pattern of CabHLH121 and CaWUS
suggests a possible feedback loop in which an enhanced
transcript level of CabHLH121 in turn represses CaWUS
transcription either directly or through TF intermediates.
A pronounced high expression level of CaWUS causes
substantial transcriptional repression of the target
CabHLH121 gene/haplotype (DBTH) in the DBTH haplo-
type-introgressed NILs (DBTH-NILCabHLH121(HAPB)) and desi
mapping parental accession (ICC 6306) with broad PW/
SWi, tall PH/SH and high SA (Figure 8a). This finding is
apparent based on a downregulation that includes a
reduced expression and accumulation of CabHLH121 tran-
scripts, thereby promoting enhanced proliferation of the
meristematic stem cell population and enlargement of the
SAM in the corresponding haplotype-introgressed NILs. By
contrast, DNDL haplotype-introgressed NILs (DNDL-NIL-
CabHLH121(HAPA)) and the desi RIL mapping parental
accession (ICC 5590) with a narrow PW/SWi, dwarf PH/SH
and less SA exhibited reduced expression of a CaWUS,
causing minimal/optimal transcriptional repression of the
target CabHLH121 gene/haplotype (DNDL), as evidenced by
its upregulation, including the enhanced expression and
accumulation of transcripts. These salient findings impli-
cate the essentiality of the effective binding and transcrip-
tional regulation (repression) of CaWUS and its target
CabHLH121 gene/haplotypes for the modulation of various
plant and SAM architectural traits (narrow versus broad
PW/SWi and dwarf versus tall PH/SH) in chickpea.
Therefore, superior transcriptional signatures (natural
alleles and haplotypes) delineated from the CabHLH121 TF
gene using a genome-wide integrated genomic strategy
are efficacious for the regulation of a desirable optimal PW
and semidwarf PH via modulating SAM morphometric
traits in chickpea. As a consequence, a CabHLH121 TF and
its DNDL and DBTH gene haplotypes regulating SAM mor-
phometric trait characteristics can be genetically manipu-
lated to achieve desirable plant architectural traits that will
be useful to develop ideal plant type cultivars restructured
with optimal PW and PH (semidwarf) in chickpea. This phe-
nomenon was evident from the NILs (DNDL-NILCabHLH121
(HAPA)) introgressed with a superior CabHLH121 gene hap-
lotype (DNDL), which exhibited an optimal PW and semid-
warf PH without any undesirable pleiotropic effects on
other agromorphological and yield component traits. The
new plant type cultivars restructured with the desirable
plant architecture of optimal PW and PH therefore showed
proficiency in enhancing yield and productivity in chickpea.
Chickpea requires a recommended planting density of 30
(row 9 row) and 10 (plant 9 plant) cm of spacing for uni-
form growth, development and maximum yield; therefore,
at most, 330 000 plants of diverse accessions can be
planted per hectare of field area for cultivation. By contrast,
we were able to accommodate 450 000 plants of the devel-
oped superior NILs (DNDL-NILCabHLH121(HAPA)) with an
© 2019 The AuthorsThe Plant Journal © 2019 John Wiley & Sons Ltd, The Plant Journal, (2019), 98, 864–883
878 Laxmi Narnoliya et al.
optimal PW (47.6 cm) and desirable PH (semidwarf with
40.5 cm PH) per hectare of cultivable field area at a planting
density of 25 (row 9 row) and 5 (plant 9 plant) cm of spac-
ing, without compromising the yield per plant and other
agronomic traits. The observed 13% higher yield/productiv-
ity in the superior DNDL-NILCabHLH121(HAPA) line was not due
to an enhanced yield per plant, however, because of their
dense plant population with a 25% increase in planting den-
sity versus that recommended for chickpea per hectare of
cultivable field area. The efficient, rapid and integrated
genomic strategy employed in this study has potential for
the development of a superior CabHLH121 gene haplotype-
introgressed desi chickpea line that produces a high yield
and productivity by restructuring the plant architectural (de-
sirable PW and PH) and SAM morphometric traits. This
achievement will essentially provide better remuneration to
farmers with rational land use, therefore ensuring global
food security in the present context of an increasing popula-
tion density and declining per capita land area.
EXPERIMENTAL PROCEDURES
Plant materials
An association panel for PW and PH was constituted by screening291 diverse chickpea germplasm accessions, including 189 desiand 102 kabuli (Table S1), from the available core collection forassociation mapping (Kujur et al., 2015a). The association panelincluded two of each cultivated desi (ICC 5590 and ICC 6306) andkabuli (ICC 12968 and ICC 16814) accessions with contrasting PWand PH traits. ICC 5590, a traditional cultivar (landrace) originatingfrom India, is a narrow and dwarf desi accession with a PW andPH of 50.6 and 38.0 cm, respectively. ICC 6306, a landrace originat-ing from the Russian Federation, is a broad and tall desi accessionwith a PW and PH of 68.1 and 66.3 cm, respectively. ICC 12968, areleased variety from India, is a narrow and semidwarf kabuliaccession with a PW and PH of 52.2 and 45.9 cm, respectively. ICC16814, a landrace originating from Portugal, is a broad and tallkabuli accession with a PW and PH of 64.3 and 57.3 cm, respec-tively. ICC 5590 and ICC 6306 were inter-crossed with each otherto develop a desi 278 F7 RIL mapping population (ICC 5590 9 ICC6306) for QTL mapping.
Field phenotyping
The germplasm accessions and mapping individuals belonging toan association panel and RIL population, respectively, were grownand phenotyped in the experimental field for two successive years(2013 and 2014) during the normal crop season as per the ran-domized complete block design (RCBD) with two replications.These procedures were performed in two geographical regions,Hyderabad (latitude 17°30N/longitude 77°20E from October toFebruary) and New Delhi (28°40N/77°20E from November toMarch), following standard agronomic practices. The PW and PHtraits were measured by estimating the mean canopy height (cm)and spread (cm), respectively, of 10 representative plants belong-ing to each accession and the RIL from the soil surface at the timeof the flower-ending and pod-setting-initiation stage. Geneticinheritance characteristics, such as the frequency distribution, CV,and broad-sense H2, of traits among accessions and RILs weredetermined according to Bajaj et al. (2015b).
Plant growth conditions
To optimize the appropriate timing for collecting the SAM andassessing their major morphometric trait variations, the seeds ofeach germplasm accession and RIL mapping individual wereplanted in pots filled with the 2:1:1 mixture of field soil: soilrite:manure, respectively. Plants were grown in the controlled growthchamber under 10-h day light and 14-h night dark conditionsbetween 20 and 22°C temperatures with 60% relative humidity.For SAM histology, 30 plants of each accession and RIL weregrown for 10–12 days after germination. SAM samples from 10-day-old seedlings (exhibiting normal and uniform growth) of eachaccession and RIL that attained full vegetative size, but not show-ing signs of any transition from the vegetative to the reproductivephase, were collected with at least three biological replicates tostudy their histology and major morphometric trait variation.
Histological assay
To perform histology including fixation, dehydration, paraplastembedding, and block preparation of SAM tissues, the shootapices of each accession and RIL were collected individually andprocessed in accordance with Gautam et al. (2016) and Ranjanet al. (2017). Tissue blocks of the appropriate size and orientationwere sectioned using a rotary microtome (Leica Biosystems, USA)to obtain strips of 8- to 10-lm-thick sections, which were flattenedby floating for 1–2 min at 50–55°C. Flattened tissue sections weremounted on charged slides and dried at 42°C overnight to attachthe tissues on the slide surface. A light microscope (Leedz MicroImaging Limited, UK) along with Zeiss AxioVision were used tocapture and visualize the SAM images. At least three independentbiological replicates with three technical replicates for each acces-sion and RIL were used for histology of SAM and the study oftheir major morphometric trait variations.
SAM morphometric measurements
The SAM architecture of each accession and RIL were measuredby seven major morphometric trait parameters, including theSWi, SH, SMWi, SH/SWi, SAL, SA, and SR using ImageJ 1.31 v(http://rsb.info.nih.gov/ij/), with three biological and technical repli-cates (Schindelin et al., 2015). SWi was measured from the notchof P1 to P2 clefts, whereas SH was estimated from the apex of theSAM to the width line at the base (Figure 1a). SMWi was calcu-lated at the midpoint of the SH, while SAL was measured by trac-ing the outer distance from the SAM apex to the P1 cleft. SA wasmeasured using the ImageJ freehand selection tool, and SR wascalculated from the emergence point of the P1 cleft to the centreof SWi (Figure 1a).
Large-scale discovery and genotyping of genome-wide
SNPs
A 96-plex GBS library was constructed using the genomic DNA of291 desi and kabuli chickpea accessions belonging to an associa-tion panel. This included two accessions representing desi (ICC5590 and ICC 6306) and kabuli (ICC 12968 and ICC 16814) with con-trasting PW and PH traits. The GBS library was sequenced with anIllumina HiSeq2000 (Illumina Inc., USA) next-generation sequenc-ing (NGS) platform, and further high-quality sequences were de-multiplexed according to Kujur et al. (2015a,b,c). The generatedsequences were mapped on the kabuli genome (Varshney et al.,2013b; http://gigadb.org/dataset/100076) following Kujur et al.(2015a,b,c). Non-erroneous SNPs were discovered and struc-turally/functionally annotated on the diverse coding and non-
© 2019 The AuthorsThe Plant Journal © 2019 John Wiley & Sons Ltd, The Plant Journal, (2019), 98, 864–883
Transcriptional tags enhance chickpea productivity 879
coding sequence components of kabuli genes and genomes (chro-mosomes/pseudomolecules and unanchored scaffolds) using cus-tomized Perl scripts, the single nucleotide polymorphism effectpredictor (SnpEff v3.1h; http://snpeff.sourceforge.net) and PFAMdatabase v27.0 (http://pfam.sanger.ac.uk) according to Kujur et al.(2015a,b,c) and Srivastava et al. (2017). The genomic distributionof SNPs in accordance with their physical positions (bp) on eightchromosome pseudomolecules of kabuli genome was visualizedwith a Circos (Krzywinski et al., 2009).
Molecular diversity and linkage disequilibrium decay
To study the molecular diversity, the genome-wide SNP genotyp-ing data generated among 291 desi and kabuli chickpea acces-sions (association panel) were analysed with PowerMarker (Liuand Muse, 2005), STRUCTURE v2.3.4 (Pritchard et al., 2000) andMEGA7 (Kumar et al., 2016). Accordingly, an unrooted neighbour-joining (NJ) phylogenetic tree was constructed among accessionsbased on Nei’s diversity coefficient (Nei et al., 1983) with 1000bootstrap replicates. The population genetic structure amongaccessions was determined according to Kujur et al. (2015a,c).
To estimate the genome-wide LD decay, the genotyping datafor SNPs physically mapped on chromosomes were analysed withPLINK (Purcell et al., 2007) and TASSEL (Bradbury et al., 2007;http://www.maizegenetics.net) following Zhao et al. (2011) andKujur et al. (2015a). The genome-wide LD decay was determinedby plotting the average r2 (frequency correlation among pair ofalleles across a pair of SNP loci) measured from the population(defined by population genetic structure) of a constituted associa-tion panel against the 50-kb uniform physical interval across eightchromosomes.
Genome-wide association study
The population structure (Q), PCA (P), and kinship (K)-matrix wereestimated using STRUCTURE, GAPIT (Lipka et al., 2012) and SPA-GeDi 1.2 (Hardy and Vekemans, 2002), respectively. To performGWAS, the SNP genotyping, phenotyping, and molecular diversityinformation generated from an association panel were analysedwith the GLM and MLM approaches of TASSEL as well as theCMLM ((P + K, K and Q + K) (Zhang et al., 2010; Kang et al.,2010)), population parameters previously determined (P3D) andthe efficient mixed model association eXpedited (EMMAX) inter-face of GAPIT (Lipka et al., 2012), according to Kujur et al. (2015a)and Upadhyaya et al. (2015). To assure the accuracy of SNP mar-ker-trait association, a quantile–quantile (Q–Q) plot-based Ben-jamini�Hochberg FDR cut-off ≤0.05 (Benjamini and Hochberg,1995) correction for multiple comparisons between observed/ex-pected �log10(P)-values and an adjusted P-value threshold of sig-nificance was employed. Based on these procedures, SNP locithat were significantly associated with the traits at a lowest FDRadjusted P-value (threshold P < 10�8) and highest R2 were identi-fied. The magnitude (R2) of PVE by the traits was measured usingan FDR-controlling model method with the SNP (adjusted P-value).
Regional association analysis and molecular haplotyping
The selected trait-associated genomic/gene regions weresequenced using the genomic DNA of 291 desi and kabuli chick-pea accessions (association panel) by employing the multiplexedamplicon sequencing protocol (per the manufacturer’s instruc-tions) of TruSeq Custom Amplicon v1.5 in the Illumina MiSeqNGS platform according to Saxena et al. (2014a). The customoligo probes targeting the selected genomic regions including
CDS (coding DNA sequences)/exons and introns as well as 3-kb ofeach URR and DRR (downstream regulatory regions) of the geneswere designed using Illumina Design Studio and synthesized fur-ther for use in the targeted multiplexed amplicon sequencing. Thepooling of amplicons (average size of 500 bp per reaction) into thecustom amplicon tubes, construction of template libraries, nor-malization of uniquely tagged pooled amplicon libraries andsequencing of generated clusters by the Illumina MiSeq platform,were performed according to Saxena et al. (2014a) and Malik et al.(2016). The mapping of high-quality amplicon sequence readsonto the kabuli chickpea genome (Varshney et al., 2013b), detec-tion of high-quality SNPs among accessions and their structural/functional annotation on the kabuli genome were performed fol-lowing Kujur et al. (2015a,c). The constitution of SNP haplotypes,determination of the SNP haplotype-based LD pattern and estima-tion of the association potential of haplotypes with the traits wereperformed according to Saxena et al. (2014a) and Kujur et al.(2015a,b).
High-resolution QTL mapping
The GBS-derived genome-wide SNP detecting polymorphismsbetween two contrasting parental accessions were genotypedusing the genomic DNA of 278 mapping individuals of a desi RILpopulation (ICC 5590 9 ICC 6306) by the Sequenom MALDI-TOFMassARRAY assay (http://www.sequenom.com) according to Sax-ena et al. (2014a,b). The Kosambi mapping function interface ofJoinMap 4.1 (www.kyazma.nl/index.php/mc.JoinMap) at a higherLOD threshold (≥4.0) was used to assign and map the SNPs intodefined linkage groups (LG1 to LG8) based on their cM geneticdistances and respective marker physical positions (bp) on thechromosomes (Kujur et al., 2015a,c). Accordingly, a high-resolu-tion genetic linkage map was constructed and visualized withMapChart v2.2 (Voorrips, 2002).
For QTL mapping, a significant CIM (composite interval map-ping) function of MapQTL 6 (van Ooijen, 2009) at a LOD cut-offscore >5.0 with 1000 permutations at a P < 0.05 was used accord-ing to Bajaj et al. (2015a,b) and Kujur et al. (2015a,b). Accordingly,the PVE (%) and additive effect (evaluated by parental origin offavourable alleles) specified by each major QTL on traits at a sig-nificant LOD were determined according to Bajaj et al. (2015a,b).The confidence interval (CI) of each significant major QTL peakwas evaluated using � 1-LOD support intervals (95% CI).
Fine-mapping and map-based cloning
A major QTL region identified by high-resolution QTL mapping(ICC 5590 9 ICC 6306) in an RIL population was targeted for fine-mapping. This procedure was performed by developing two BC4F4NILs according to the strategies depicted in Figure S6. One BC4F4NIL was developed as a DNDL (desi narrow PW/SWi, dwarf PH/SH,and low SA)-NILCaqPW/PH/SWi/SH/SA8.1. Another BC4F4 NIL was devel-oped as a DBTH (desi broad PW/SWi, tall PH/SH, and high SA)-NILCaqPW/PH/SWi/SH/SA8.1. The generated NILs were inter-crossedamong each other to develop a F2 mapping population (DNDL-NIL 9 DBTH-NIL) consisting of 168 individuals. SNPs showing par-ental polymorphisms between ICC 5590 and ICC 6306 at this QTLgenomic interval were genotyped in 168 mapping individualsusing the MALDI-TOF assay according to Saxena et al. (2014a,b).Field and controlled environment phenotyping of mapping indi-viduals for the traits and QTL mapping were performed accordingto the aforesaid strategies. For the QTL region-specific associationanalysis, a target QTL genomic interval was sequenced using thegenomic DNA of the parental accessions (ICC 5590 and ICC 6306)and two selected homozygous individuals of a mapping
© 2019 The AuthorsThe Plant Journal © 2019 John Wiley & Sons Ltd, The Plant Journal, (2019), 98, 864–883
880 Laxmi Narnoliya et al.
population (DNDL-NIL 9 DBTH-NIL) that contrasted with the stud-ied traits. The sequencing and mapping of high-quality ampliconsequence reads as well as the detection of SNPs and their struc-tural/functional annotation onto the kabuli genome were per-formed following Saxena et al. (2014a,b), Kujur et al. (2015a,c)and Malik et al. (2016). The genotyping of SNPs mined at a majorQTL genomic interval among accessions, phenotyping and traitassociation analysis were performed following the aforesaidmethods. For progeny analysis, homozygous recombinant andhomozygous non-recombinant individuals derived from the NILswith contrasting traits were selected based on their genetic consti-tution and considering the recombination among SNPs flanking/tightly linked to a major QTL, as well as a strong trait-associatedgene and its constituted DNDL and DBTH haplotypes. The selectedrecombinant and non-recombinant progenies were phenotypedfor the studied traits. Significant variations of traits betweenselected recombinant and non-recombinant progenies were evalu-ated by a statistical one-tailed t-test.
For marker (haplotype)-assisted foreground selection, SNPsflanking/tightly linked to a major QTL as well as a strong trait-associated gene and its DNDL and DBTH haplotypes were geno-typed among individuals of the back-cross mapping populationusing the MALDI-TOF assay following Saxena et al. (2014a,b).For marker (haplotype)-assisted background selection, 1536 SNPsshowing polymorphisms between parental accessions (ICC 5590and ICC 6306) of an RIL mapping population (ICC 5590 9 ICC6306) and that mapped uniformly across the chromosomes ofthe kabuli genome were genotyped in the selected back-crossmapping individuals by the Illumina GoldenGate SNP genotyp-ing assay according to Bajaj et al. (2015a) (Figure S6). Large-scale mapping individuals of the back-cross population werephenotyped for the traits in the field as well as in the controlledgrowth conditions with at least three replications, according tothe aforementioned methods. The NILs introgressed with theDNDL and DBTH haplotypes were further phenotyped for theSWi, SH and SA traits following the described histologicalassay.
Scanning electron microscopy
For the SEM assay, 10-day-old SAM tissues of RIL mapping paren-tal accessions (ICC 5590 and ICC 6306) as well as DNDL and DBTHgene haplotype-introgressed NILs were fixed in FAA (3.7%formaldehyde, 50% ethanol and 5% acetic acid) solution and trea-ted overnight with 0.2% osmium tetroxide followed by serial etha-nol dehydration according to Kujur et al. (2015a) and Ranjan et al.(2017). The dehydrated samples were further processed by criti-cal-point drying and examined by SEM (Evo LS25; Zeiss). All theSEM-captured images of SAM were processed with ImageJ tomeasure their major morphometric parameters and variationsamong accessions and NILs.
Differential expression profiling
For in silico global expression profiling of the trait-associatedgenes, the raw FASTAQ transcriptome sequence data of germinat-ing seedlings, young leaves, SAM, and eight different develop-ment stages of flower buds/flowers of a desi chickpea accessionICC 4958 were acquired from the NCBI-SRA (sequence readarchive) database (Singh et al., 2013), and the high-qualitysequences were mapped on the kabuli genome. Global transcrip-tome profiling was performed according to Srivastava et al.(2016). The expression profiles of trait-associated chickpea geneswere retrieved, analysed and visualized with a heat map using theMultiExperiment Viewer (MeV, http://www.tm4.org/mev).
For experimental validation, RIL mapping parental accessions(ICC 5590 and ICC 6306) and kabuli accessions (ICC 12968 and ICC16814), as well as DNDL and DBTH gene haplotype-introgressedNILs, were grown in the controlled environment according to theaforementioned strategy. High-quality RNA was isolated from theshoot, leaf and SAM tissues of 10-day-grown plants of thedescribed accessions/NILs with biological and technical replicates.Quantitative real-time (RT)-PCR was performed by amplifying theisolated RNA with specific primers designed based on the gene/haplotype and an internal control gene, Actin, following Bajaj et al.(2015b) and Upadhyaya et al. (2015). The 2�ΔΔCT method and two-tailed t-test were employed to determine the expression pattern ofthe gene/haplotypes and their significant differences in expression,respectively, among accessions/NILs (Bajaj et al., 2015b;Mukhopadhyay and Tyagi, 2015). The expression profile was visu-alized with a heat map using MeV (http://www.tm4.org/mev).
Identification and analysis of cis-regulatory elements
The sequence of the 3-kb URR of a strong trait-associated genewas acquired based on the respective genomic coordinates andstructural annotation of the kabuli genome. The obtained URRsequence was further analysed in PlantPAN 2.0 (http://plantpan2.itps.ncku.edu.tw/) to scan for the presence of cis-regulatory ele-ments or known TF-binding sites. These known TF-binding sitesof a trait-associated gene were correlated with the SNPs and hap-lotypes identified based on the URR of a target gene among 291desi and kabuli chickpea accessions (association panel), as well asthe aforesaid in silico and experimental differential expressionprofile of the gene/haplotypes.
ChIP (chromatin immunoprecipitation) assay
For the ChIP analysis, the SAM of 10-day-old plants of individual NILsintrogressed with DNDL and DBTH gene haplotypes were used toisolate nuclei to immunoprecipitate the sonicated chromatin suspen-sion with the protein�DNA complex using a TF gene-specific anti-body (Roche, USA) as previously described (Sun et al., 2015; Tianet al., 2017). The antibody was specific to a TF gene that was pre-dicted to bind in the TF-binding site (cis-regulatory element region)of the URR of a strong trait-associated gene. Three biological/techni-cal replicates representing individual haplotypes along with immu-noglobin G (IgG) as a control were assayed in the ChIP experiment.For ChIP sequencing (ChIP-seq), libraries were prepared from chro-matin immunoprecipitated DNA and sequenced using an IlluminaHiSeq4000 (Illumina Inc., USA) NGS platform according to the manu-facturer’s instructions. Quality checking of the sequences and map-ping of high-quality sequences on the kabuli genome wereperformed according to Kujur et al. (2015a,c). The ChIP-enrichedpeaks were identified based on the sequence coverage at individualphysical positions (bp) on the kabuli genome (Varshney et al., 2013b)following Sun et al. (2015). The ChIP-quantitative PCR (ChIP-qPCR)assay was performed by amplifying the equally quantified said chro-matin immunoprecipitated DNA isolated from the NILs with gene/haplotype-specific primers. Primer pairs were designed targeting theChIP-seq validated TF-binding sites identified in the URR of a strongtrait-associated gene following the aforementioned RT-PCR meth-ods. The outcomes of both ChIP-seq and ChIP-PCR assays wereascertained by three independent sequencing and PCR experimentsusing three biological/technical replicates.
ACKNOWLEDGEMENTS
We are very much grateful to Mr Sube Singh, lead scientific offi-cer, Grain Legumes Research Program/GenBank, ICRISAT, Hyder-abad for assisting in collecting the multienvironments field
© 2019 The AuthorsThe Plant Journal © 2019 John Wiley & Sons Ltd, The Plant Journal, (2019), 98, 864–883
Transcriptional tags enhance chickpea productivity 881
phenotyping data for the germplasm accessions and mappingpopulation. Constant assistance was provided by all the scientificand technical staff from NIPGR and IARI, New Delhi, India andICRISAT, Hyderabad to conduct this research study is acknowl-edged. We are also thankful to the central instrumentation facility(CIF), the plant growth facility (PGF) and the DBT-eLibrary Consor-tium (DeLCON) of NIPGR, New Delhi for providing timely supportand access to e-resources for this research work.
FUNDING
The financial support for this study was provided by a
research grant from the Department of Biotechnology
(DBT), Government of India. LN, UB, VT, and AD acknowl-
edge the UGC (University Grants Commission) and DBT,
India for research fellowship awards.
CONFLICT OF INTEREST
The authors declare that they have no conflicts of interest.
AUTHORS CONTRIBUTIONS
LN, UB, DB, NM, and VT performed the field and labora-
tory experiments as well as drafted the manuscript. ST,
VSH, and HDU helped in constitution/generation of associ-
ation panel and mapping populations and performed field
phenotyping. AD, DB, and AS helped in the large-scale
SNP genotyping data analyses. AKS, AKT, and SKP con-
ceived and designed the research study, guided data analy-
sis and interpretation, participated in drafting and
correcting the manuscript critically. All authors gave the
final approval of the version to be published.
SUPPORTING INFORMATION
Additional Supporting Information may be found in the online ver-sion of this article.Figure S1. Genomic distribution and structural annotation of333001 SNPs physically mapped on eight chromosomes andunanchored scaffolds of kabuli genome.
Figure S2. Determination of molecular diversity, phylogenetic rela-tionship and historical recombination (linkage disequilibrium)among 291 desi (189) and kabuli (102) germplasm accessionsbelonging to an association panel of chickpea.
Figure S3. Construction of a high-density genetic linkage map andmolecular mapping of major QTLs governing vital plant architec-tural and SAM morphometric traits in chickpea.
Figure S4. Frequency distribution and boxplots depicting the phe-notypic variation of major plant architectural and SAM morpho-metric traits in a RIL mapping population (ICC 5590 9 ICC 6306).
Figure S5. A hierarchical cluster display depicting phenotypic cor-relation among five major plant architectural and SAM morpho-metric traits in 278 mapping individuals of a RIL population (ICC5590 9 ICC 6306).
Figure S6. Schematic illustration of the comprehensive strategiesfollowed to develop NILs of chickpea by marker (haplotype)-assisted foreground and background selection.
Figure S7. Relative proportion of known cis-regulatory elements/transcription factor (TF) binding sites predicted in a 3 kb regula-tory (promoter) sequence upstream to the translation initiationcodon (ATG) of CabHLH121.
Table S1. Germplasm accessions utilized for dissection of majorplant architectural and SAM morphometric traits in chickpea.
Table S2. Genetic inheritance pattern of traits in the associationpanel and RIL mapping population used for GWAS and QTL map-ping.
Table S3. Genome-wide SNPs mined for GWAS of major plantarchitectural and SAM morphometric traits in chickpea.
Table S4. SNPs employed to construct a high-density genetic link-age map (ICC 5590 9 ICC 6306) of chickpea.
Table S5. Genome-wide SNPs mapped on eight chromosomes ofa chickpea genetic linkage map.
Table S6. Molecular mapping of major QTLs governing five majorplant architectural and SAM morphometric traits in chickpea.
Table S7. SNPs mined within a major QTL interval for QTL region-specific association analysis in chickpea.
Table S8. Comparative assessment of agronomic trait characteris-tics between haplotype-introgressed NILs vis-�a-vis mapping par-ental accessions and a high-yielding Indian variety of chickpea.
Data S1. Results.
REFERENCES
Bajaj, D., Upadhyaya, H.D., Khan, Y. et al. (2015a) A combinatorial approach
of comprehensive QTL-based comparative genome mapping and tran-
script profiling identified a seed weight-regulating candidate gene in
chickpea. Sci. Rep. 5, 9264.
Bajaj, D., Saxena, M.S., Kujur, A. et al. (2015b) Genome-wide conserved
non-coding microsatellite (CNMS) marker-based integrative genetical
genomics for quantitative dissection of seed weight in chickpea. J. Exp.
Bot. 66, 1271–1290.Barton, M.K. (2010) Twenty years on: the inner workings of the shoot apical
meristem, a developmental dynamo. Dev. Biol. 341, 95–113.Benjamini, Y. and Hochberg, Y. (1995) Controlling the false discovery rate: a
practical and powerful approach to multiple testing. J. R. Stat. Soc. Ser-
ies B Stat. Methodol. 57, 289–300.Bradbury, P.J., Zhang, Z., Kroon, D.E., Casstevens, T.M., Ramdoss, Y. and
Buckler, E.S. (2007) TASSEL: Software for association mapping of com-
plex traits in diverse samples. Bioinformatics, 23, 2633–2635.Buechel, S., Leibfried, A., To, J.P., Zhao, Z., Andersen, S.U., Kieber, J.J. and
Lohmann, J.U. (2010) Role of A-type ARABIDOPSIS RESPONSE REGULA-
TORS in meristem maintenance and regeneration. Eur. J. Cell Biol. 89,
279–284.Cai, G., Yang, Q., Chen, H., Yang, Q., Zhang, C., Fan, C. and Zhou, Y. (2016)
Genetic dissection of plant architecture and yield-related traits in Bras-
sica napus. Sci. Rep. 6, 21625.
Evenson, R.E. and Gollin, D. (2003) Assessing the impact of the green revo-
lution, 1960 to 2000. Science, 300, 758–762.Gautam, V., Singh, A., Singh, S. and Sarkar, A.K. (2016) An efficient LCM-
based method for tissue specific expression analysis of genes and miR-
NAs. Sci. Rep. 6, 21577.
Gupta, S., Nawaz, K., Parween, S., Roy, R., Sahu, K., Kumar Pole, A., Khan-
dal, H., Srivastava, R., Kumar Parida, S. and Chattopadhyay, D. (2016)
Draft genome sequence of Cicer reticulatum L., the wild progenitor of
chickpea provides a resource for agronomic trait improvement. DNA
Res. 24, 1–10.Hardy, O.J. and Vekemans, X. (2002) SPAGeDi: a versatile computer pro-
gram to analyse spatial genetic structure at the individual or population
levels. Mol. Ecol. Notes, 2, 618–620.Jain, M., Misra, G., Patel, R.K. et al. (2013) A draft genome sequence of the
pulse crop chickpea (Cicer arietinum L.). Plant J. 74, 715–729.Jin, J., Huang, W., Gao, J.P., Yang, J., Shi, M., Zhu, M.Z., Luo, D. and Lin,
H.X. (2008) Genetic control of rice plant architecture under domestica-
tion. Nat. Genet. 40, 1365–1369.Kang, H.M., Sul, J.H., Service, S.K., Zaitlen, N.A., Kong, S.Y., Freimer, N.B.,
Sabatti, C. and Eskin, E. (2010) Variance component model to account
for sample structure in genome-wide association studies. Nat. Genet. 42,
348–354.
© 2019 The AuthorsThe Plant Journal © 2019 John Wiley & Sons Ltd, The Plant Journal, (2019), 98, 864–883
882 Laxmi Narnoliya et al.
Krzywinski, M., Schein, J., Birol, I., Connors, J., Gascoyne, R., Horsman, D.,
Jones, S.J. and Marra, M.A. (2009) Circos: an information aesthetic for
comparative genomics. Genome Res. 19, 1639–1645.Kujur, A., Bajaj, D., Upadhyaya, H.D. et al. (2015a) A genome-wide SNP scan
accelerates trait-regulatory genomic loci identification in chickpea. Sci.
Rep. 5, 11166.
Kujur, A., Upadhyaya, H.D., Shree, T. et al. (2015b) Ultra-high density intra-
specific genetic linkage maps accelerate identification of functionally rel-
evant molecular tags governing important agronomic traits in chickpea.
Sci. Rep. 5, 9468.
Kujur, A., Bajaj, D., Upadhyaya, H.D. et al. (2015c) Employing genome-wide
SNP discovery and genotyping strategy to extrapolate the natural allelic
diversity and domestication patterns in chickpea. Front. Plant Sci. 6, 162.
Kujur, A., Upadhyaya, H.D., Bajaj, D., Gowda, C.L., Sharma, S., Tyagi, A.K. and
Parida, S.K. (2016) Identification of candidate genes and natural allelic vari-
ants for QTLs governing plant height in chickpea. Sci. Rep. 6, 27968.
Kumar, J., Choudhary, A.K., Solanki, R.K. and Pratap, A. (2011) Towards
marker-assisted selection in pulses: a review. Plant Breed. 130, 297–313.Kumar, S., Stecher, G. and Tamura, K. (2016) MEGA7: Molecular evolution-
ary genetics analysis version 7.0 for bigger datasets. Mol. Biol. Evol. 33,
1870–1874.Laux, T., Mayer, K.F., Berger, J. and J€urgens, G. (1996) The WUSCHEL gene
is required for shoot and floral meristem integrity in Arabidopsis. Devel-
opment, 122, 87–96.Lee, D.K., Parrott, D.L., Adhikari, E., Fraser, N. and Sieburth, L.E. (2016) The
mobile bypass signal arrests shoot growth by disrupting shoot apical
meristem maintenance, cytokinin signaling, and WUS transcription fac-
tor expression. Plant Physiol. 171, 2178–2190.Leibfried, A., To, J.P., Busch, W., Stehling, S., Kehle, A., Demar, M., Kieber, J.J.
and Lohmann, J.U. (2005) WUSCHEL controls meristem function by direct
regulation of cytokinin-inducible response regulators.Nature, 438, 1172–1175.Leiboff, S., Li, X., Hu, H.C. et al. (2015) Genetic control of morphometric
diversity in the maize shoot apical meristem. Nat. Commun. 6, 8974.
Lipka, A.E., Tian, F., Wang, Q., Peiffer, J., Li, M., Bradbury, P.J., Gore, M.A.,
Buckler, E.S. and Zhang, Z. (2012) GAPIT: genome association and pre-
diction integrated tool. Bioinformatics, 2, 2397–2399.Liu, K. and Muse, S.V. (2005) PowerMarker: an integrated analysis environ-
ment for genetic marker analysis. Bioinformatics, 21, 2128–2129.Malik, N., Dwivedi, N., Singh, A.K., Parida, S.K., Agarwal, P., Thakur, J.K. and
Tyagi, A.K. (2016) An integrated genomic strategy delineates candidate
mediator genes regulating grain size and weight in rice. Sci. Rep. 6, 23253.
Mukhopadhyay, P. and Tyagi, A.K. (2015) OsTCP19 influences developmen-
tal and abiotic stress signaling by modulating ABI4-mediated pathways.
Sci. Rep. 5, 9998.
Nei, M., Tajima, F. and Tateno, Y. (1983) Accuracy of estimated phylogenetic
trees frommolecular data. II. Gene frequency data. J. Mol. Evol. 19, 153–170.van, Ooijen, J.W. (2009) MapQTL 6, Software for the Mapping of Quantita-
tive Trait Loci in Experimental Populations of Diploid Species. Wagenin-
gen, Netherlands: Kyazma BV.
Parween, S., Nawaz, K., Roy, R. et al. (2015) An advanced draft genome
assembly of a desi type chickpea (Cicer arietinum L.). Sci. Rep. 5, 12806.
Peng, J., Richards, D.E., Hartley, N.M. et al. (1999) Green revolution genes
encode mutant gibberellin response modulators. Nature, 400, 256–261.Penmetsa, V.R., Carrasquilla-Garcia, N., Bergmann, E.M. et al. (2016) Multiple
post-domestication origins of kabuli chickpea through allelic variation in a
diversification-associated transcription factor. New Phytol. 211, 1440–1451.Pickersgill, B. (2007) Domestication of plants in the Americas: insights from
Mendelian and molecular genetics. Ann. Bot. 100, 925–940.Pritchard, J.K., Stephens, M. and Donnelly, P. (2000) Inference of population
structure using multilocus genotype data. Genetics, 155, 945–959.Purcell, S., Neale, B., Todd-Brown, K. et al. (2007) PLINK: a tool set for
whole-genome association and population-based linkage analyses. Am.
J. Hum. Genet. 81, 559–575.Purushothaman, R., Upadhyayaa, H.D., Gaur, P.M., Gowda, C.L.L. and Krish-
namurthya, L. (2014) Kabuli and desi chickpeas differ in their require-
ment for reproductive duration. Field Crops Res. 163, 24–31.Ranjan, R., Khurana, R., Malik, N., Badoni, S., Parida, S.K., Kapoor, S. and Tyagi,
A.K. (2017) bHLH142 regulates various metabolic pathway-related genes to
affect pollen development and anther dehiscence in rice. Sci. Rep. 7, 43397.
Reinhardt, D. and Kuhlemeier, C. (2002) Plant architecture. EMBO Rep. 3,
846–851.
Sakamoto, T., Miura, K., Itoh, H. et al. (2004) An overview of gibberellin
metabolism enzyme genes and their related mutants in rice. Plant Phys-
iol. 134, 1642–1653.Saxena, M.S., Bajaj, D., Das, S., Kujur, A., Kumar, V., Singh, M., Bansal,
K.C., Tyagi, A.K. and Parida, S.K. (2014a) An integrated genomic
approach for rapid delineation of candidate genes regulating agro-mor-
phological traits in chickpea. DNA Res. 21, 695–710.Saxena, M.S., Bajaj, D., Kujur, A., Das, S., Badoni, S., Kumar, V., Singh, M.,
Bansal, K.C., Tyagi, A.K. and Parida, S.K. (2014b) Natural allelic diversity,
genetic structure and linkage disequilibrium pattern in wild chickpea.
PLoS ONE, 9, e107484.
Saxena, R.K., Obala, J., Sinjushin, A., Kumar, C.V.S., Saxena, K.B. and
Varshney, R.K. (2017) Characterization and mapping of Dt1 locus which
co-segregates with CcTFL1 for growth habit in pigeonpea. Theor. Appl.
Genet. 130, 1773–1784.Schindelin, J., Rueden, C.T., Hiner, M.C. and Eliceiri, K.W. (2015) The Ima-
geJ ecosystem: an open platform for biomedical image analysis. Mol.
Reprod. Dev. 82, 518–529.Schuster, C., Gaillochet, C., Medzihradszky, A., Busch, W., Daum, G., Krebs,
M., Kehle, A. and Lohmann, J.U. (2014) A regulatory framework for shoot
stem cell control integrating metabolic, transcriptional, and phytohor-
mone signals. Dev. Cell, 28, 438–449.Singh, V.K., Garg, R. and Jain, M. (2013) A global view of transcriptome
dynamics during flower development in chickpea by deep sequencing.
Plant Biotechnol. J. 11, 691–701.Sparks, E.E. and Benfey, P.N. (2014) HEC of a job regulating stem cells. Dev.
Cell, 28, 349–350.Srivastava, R., Bajaj, D., Malik, A., Singh, M. and Parida, S.K. (2016) Tran-
scriptome landscape of perennial wild Cicer microphyllum uncovers
functionally relevant molecular tags regulating agronomic traits in chick-
pea. Sci. Rep. 6, 33616.
Srivastava, R., Upadhyaya, H.D., Kumar, R., Daware, A., Basu, U., Shimray,
P.W., Tripathi, S., Bharadwaj, C., Tyagi, A.K. and Parida, S.K. (2017) A
multiple QTL-Seq strategy delineates potential genomic loci governing
flowering time in chickpea. Front. Plant Sci. 8, 1105.
Sun, T., Zhang, Y., Li, Y., Zhang, Q., Ding, Y. and Zhang, Y. (2015) ChIP-seq
reveals broad roles of SARD1 and CBP60 g in regulating plant immunity.
Nat. Commun. 6, 10159.
Thompson, A.M., Crants, J., Schnable, P.S., Yu, J., Timmermans, M.C.,
Springer, N.M., Scanlon, M.J. and Muehlbauer, G.J. (2014) Genetic con-
trol of maize shoot apical meristem architecture. G3, 4, 1327–1337.Thompson, A.M., Yu, J., Timmermans, M.C., Schnable, P., Crants, J.C., Scanlon,
M.J. and Muehlbauer, G.J. (2015) Diversity of maize shoot apical meristem
architecture and its relationship to plant morphology.G3, 5, 819–827.Tian, H., Wang, X., Guo, H., Cheng, Y., Hou, C., Chen, J.G. and Wang, S.
(2017) NTL8 regulates trichome formation in Arabidopsis by directly acti-
vating R3 MYB genes TRY and TCL1. Plant Physiol. 174, 2363–2375.Upadhyaya, H.D., Bajaj, D., Das, S. et al. (2015) A genome-scale integrated
approach aids in genetic dissection of complex flowering time trait in
chickpea. Plant Mol. Biol. 89, 403–420.Varshney, R.K., Mohan, S.M., Gaur, P.M. et al. (2013a) Achievements and
prospects of genomics-assisted breeding in three legume crops of the
semiarid tropics. Biotech. Adv. 31, 1120–1134.Varshney, R.K., Song, C., Saxena, R.K. et al. (2013b) Draft genome sequence
of chickpea (Cicer arietinum) provides a resource for trait improvement.
Nat. Biotechnol. 31, 240–246.Voorrips, R.E. (2002) MapChart: Software for the graphical presentation of
linkage maps and QTLs. J. Hered. 93, 77–78.Wang, Y. and Li, J. (2008) Molecular basis of plant architecture. Annu. Rev.
Plant Biol. 59, 253–279.Yang, X.C. and Hwa, C.M. (2008) Genetic modification of plant architecture
and variety improvement in rice. Heredity, 101, 396–404.Yano, K., Yamamoto, E., Aya, K. et al. (2016) Genome-wide association
study using whole-genome sequencing rapidly identifies new genes
influencing agronomic traits in rice. Nat. Genet. 48, 927–934.Zhang, Z., Ersoz, E., Lai, C.Q. et al. (2010) Mixed linear model approach
adapted for genome-wide association studies. Nat. Genet. 42, 355–360.Zhao, K., Tung, C.W., Eizenga, G.C. et al. (2011) Genome-wide association
mapping reveals a rich genetic architecture of complex traits in Oryza
sativa. Nat. Commun. 2, 467.
© 2019 The AuthorsThe Plant Journal © 2019 John Wiley & Sons Ltd, The Plant Journal, (2019), 98, 864–883
Transcriptional tags enhance chickpea productivity 883
Top Related